Development of a Bio-robotic Swimmer Based on the California Sea Lion

Development of a Bio-robotic Swimmer Based on the California Sea Lion | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Development of a Bio-robotic Swimmer Based on the California Sea Lion Nicholas Marcouiller, Shraman Kadapa, Anthony Drago, Frank Fish, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7455024/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The development of unmanned underwater vehicles (UUVs) capable of operating in complex environments—such as coastal regions with obstacles and dynamic flows—requires new and effective maneuvering techniques with high agility to overcome the limitations of current underwater systems. UUVs that can operate in these zones have broad applications, including environmental monitoring, defense, and infrastructure inspection. By studying the swimming and maneuvering strategies of marine organisms, researchers can develop UUVs that integrate biologically inspired characteristics to enhance performance. The California sea lion ( Zalophus californianus ) was selected as a biological model due to its swimming and maneuvering capabilities in both the open ocean and through the high-energy surf zone. This paper presents the development of a novel, multi-bodied, bio-robotic system with flipper-based propulsion modeled after the California sea lion. An articulatable head and pelvis, flexible fore flippers that generate 3D forces, and adjustable hind flippers were identified as potential contributors to its mobility, as supported by existing research and video analysis. The system serves as a research platform for systematically evaluating how these features influence swimming and maneuvering. Experimental results demonstrate the system's ability to use hydrostatic and hydrodynamic forces to move repeatably in 3D space, providing a foundation for assessing the role of body articulation and flipper movements in underwater locomotion. Biological sciences/Ecology Earth and environmental sciences/Ecology Physical sciences/Engineering Earth and environmental sciences/Ocean sciences Underwater Robot multi-body Sea lion Maneuverable Bio-inspired Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 Figure 18 Figure 19 Figure 20 Figure 21 Figure 22 Figure 23 Figure 24 Figure 25 1 Introduction Unmanned Underwater Vehicles (UUVs) play a crucial role across a wide array of maritime operations, including scientific research, commercial endeavors, and defense applications; however, their limited maneuverability often restricts their operational reach. While UUVs are increasingly used in the open ocean, there is growing interest in expanding their capabilities to littoral zones, rivers, harbors, and surf zones where maneuverability plays a key role in their success [1], [2]. These missions would frequently require UUVs to navigate tight spaces, avoid obstacles, and execute rapid turns and directional changes in response to turbulent and occluded flow conditions. To operate effectively in these complex, high-energy environments, UUVs would need to be highly agile and maneuverable, requiring further advancements of UUVs. Aquatic animals, particularly those adept at navigating high-energy environments, serve as valuable models for enhancing the maneuverability of underwater vehicles. Their remarkable locomotion strategies, which include coordinated body segment movement and adaptive high-speed swimming patterns, enable agile maneuvering in complex, cluttered environments [3]. Studying the morphology, physiology, and kinematics of these biological systems directly informs the development of robotic platforms with improved propulsive efficiency, maneuverability, agility, stability, station holding, and stealth [4]. While significant progress has been made in understanding underwater animal locomotion [5], [6], current underwater systems have yet to fully exploit the characteristics such as body-fin coordination, adaptable compliance, or dynamic vortex manipulation that enable certain aquatic animals to maneuver effectively in diverse flow conditions [7], [8], [9]. Among these, the California sea lion ( Zalophus californianus ) stands out as a prime subject of advanced research [10], [11], [12], [13]. The California sea lion is capable of swimming at 9.7 m/s and diving to a maximum depth of 274 m [14], [15], [16]. Its exceptional ability to navigate complex underwater environments, where it encounters obstacles and energetic flows, makes it an invaluable model for a bio-inspired design [17], [18]. Notably, sea lions have achieved turning rates of up to 690°/s and minimum unpowered (i.e., without active propulsion) turning radii as small as 0.09 body lengths [18]. This high degree of maneuverability and agility can be partly attributed to the animal's coordinated use of its powerful flippers and flexible body [10], [17]. Ultimately, understanding these animals' swimming and maneuvering strategies provides a crucial foundation for applying such techniques in dynamic conditions, offering key insights for developing UUVs capable of operating across varied flow regimes. The objective of this work was to develop a freely swimming bio-robotic research platform, modeled after the California sea lion, to explore morphological and kinematic traits that contribute to effective swimming and maneuvering. This platform, named the Stroke Experimentation and Maneuver Optimizing Underwater Robot (SEAMOUR) (Figure 1), was specifically designed to investigate various propulsion and maneuvering techniques through coordinated flipper and body motions. Initial tests explored rectilinear swimming approaches and the use of multiple control surfaces—located fore and aft of the robot's center of mass—to produce pitch and yaw turns. Insights from these studies enabled SEAMOUR to execute freely swimming, coordinated multi-axis maneuvers using independently actuated control surfaces distributed along its body. The system's repeatable and versatile functionality makes it ideal for understanding the diverse swimming and maneuvering techniques employed by California sea lions. A complementary numerical model of the bio-robotic platform was also developed to evaluate and explore various swimming strategies; however, its details are presented in a separate study [19] and are not the focus of this paper. Studies using both the physical and numerical systems provide a foundation for identifying design features that can enhance UUV maneuverability, with future work exploring performance in dynamic flow environments. While traditional UUVs possess varied strengths, their inherent characteristics consistently limit their effectiveness in complex, high-energy environments. For example, torpedo-shaped UUVs that use a single axial propeller operate with limited control surfaces and are designed to prioritize stability and speed for navigating unobstructed open water, an environment where they consistently perform well [20]. While box-frame UUVs can be equipped with a variety of payloads and have multiple propulsors, their large, flat shape generates significant drag at high speeds and can be easily perturbed in energetic environments [21]. Lastly, glider-type UUVs are highly efficient for long-distance travel in open water but have limited speed and agility [22]. As a result, there remains a critical need for systems capable of operating in dynamic coastal environments. Biomimetic robotics has demonstrated significant success in developing versatile systems that leverage intricate biological traits, offering a powerful alternative to conventional engineering paradigms. While propeller-based systems allow for precise and predictable adjustments to speed and direction through direct control of rotation speed and thrust, animals, in contrast, use intricate movements of their flexible appendages and bodies to generate multidimensional forces [23], [18]. Over the years, numerous researchers have worked on developing systems inspired by marine creatures, including various fish to understand the use of multiple fins for propulsion and maneuvering [24], [25], [26]; sea turtles and cownose rays to explore the potential of soft actuators [27], [28]; and sea snakes/eels to investigate the use of multi-bodied systems [29], [30], to name a few. By incorporating biologically derived characteristics into traditionally shaped UUVs, some researchers have already demonstrated improvements in system maneuverability [31], [32]. The application of biological principles consistently underscores the immense potential of biomimetics, further validating the approach of using highly agile models like the California sea lion to significantly advance UUV maneuverability and performance across diverse environments. The remainder of this paper is organized as follows. The first section provides a description of common techniques used by sea lions for swimming and maneuvering, followed by the extraction of flipper and body kinematics. The next section describes the design of the bio-robotic system, with the following section describing the experimental setup and data collection techniques. A combined results and discussion section examine the static stability of the system’s design, rectilinear swimming approaches, pitch and yaw turns, and coordinated maneuvers. Finally, the paper will conclude with a summary of key findings of the work and a discussion of future research directions. 2 Background of California Sea Lion Swimming and Maneuvering The California sea lion ( Zalophus californianus ) was selected as the model organism for this study due to its versatility and exceptional performance in both open ocean and dynamic coastal environments. Sea lions regularly navigate complex habitats including kelp forests, rocky shorelines, and turbulent surf zones, making maneuverability a critical factor in their ecological success [ 18 ]. Adult sea lions, which typically measure around 1.81 m in length, exhibit impressive aquatic performance—cruising at sustained speeds of 3.5 m/s [ 17 ], executing porpoising leaps of 0.94 m in height and 1.90 m in length at 2 m/s, and performing agile banking turns with minimum radii of 0.29 m at 3.2 m/s [ 18 ], [ 33 ]. Their capabilities also extend beyond the water, as they can transition to land and move efficiently using walking or galloping gaits [ 10 ], [ 34 ]. Sea lions retain mobility in both the fore and hind limbs, providing distinct mechanical strategies for propulsion, stabilization, and maneuvering. In the water, the fore flippers serve as the primary propulsors and control surfaces, while the hind flippers assist with stabilization and maneuvering [ 18 ], [ 33 ]. Together, these traits make the California sea lion an ideal biological model for investigating aquatic locomotion, offering high maneuverability through coordinated use of its flippers and body across a range of aquatic environments (Fig. 2 ). The California sea lion primarily relies on its fore flippers for propulsion during swimming, with minimal contribution from the hind flippers or other body sections. The propulsive stroke has been categorized into three distinct phases: recovery, power, and paddle (Fig. 3 ) [ 12 ], defined by the motion at the base of the flipper (shoulder/elbow joints). The recovery phase begins and ends with the fore flippers retracted, medially rotated, and held against the ventrolateral surface of the body. During the recovery phase, the flippers are laterally rotated and abducted, bringing them near a zero angle of attack with the flow. As the flippers are further abducted, they rotate to a positive angle of attack. Lateral rotation slows as medial rotation begins. The power phase starts as medial rotation continues, and the flippers are adducted, resulting in a noticeable dorsal flexion of the distal two-thirds. The flippers' orientation then changes from a positive to a negative angle of attack as they pass the body midline. Rotation slows as adduction continues until the flippers are fully extended beneath the body. At the start of the paddle phase of the stroke, the flippers are maximally adducted and medial rotation continues as they are retracted. The flippers are then forcefully brought upward and inward toward the body [ 12 ]. Finally, the flippers return to a streamlined position next to the body [ 35 ]. In steady swimming, the hind flippers are typically positioned in an inverted V-shape with little head movement [ 17 ] and between strokes the fore flippers may be adducted against the body to create a streamlined profile [ 33 ]. Throughout these studies this stroke is referred to as the characteristic stroke and acts as the foundation to other swimming approaches that are explored using the bio-robotic system. The California sea lion employs a variety of strategies to maneuver through the water, with one of the most common being the execution of a banking turn. This maneuver showcases the animal’s high maneuverability and highlights the coordinated use of body and flipper movements required to rotate the body about multiple axes. Notably, banking turns enable sharp, unpowered changes in direction, achieving smaller turning radii than powered turns [ 18 ], [ 33 ]. To initiate a banking turn, the sea lion first rotates its head laterally toward the desired direction of travel. This motion is followed by abduction of the fore flippers and axial body rotation about the longitudinal axis. As the body continues to roll, the hind flippers are angled dorsally into the vertical plane, with an abduction of the digits. Simultaneously, the body undergoes lateral flexion as the head and pelvic region bend toward the center of the turn as the body is bent dorsally producing a continuous, coordinated turn. To complete the turn the animal straightens its body, reducing flexion of the vertebral column and re-aligning with the intended direction of travel. At the same time, the fore flippers are used to accelerate the body out of the turn as the flippers are adducted and retracted to lay appressed against the body. The body then reverses its roll, returning to its starting orientation to resume steady swimming or gliding (Fig. 5 ) [ 17 ], [ 33 ]. The coordinated body and flipper kinematics demonstrated in this maneuver provide valuable biomechanical principles for informing the design of the robotic system and guiding the development of maneuverability strategies. 3 Kinematics of Biological System The kinematics of flipper and body motions during swimming and maneuvering were identified from videos of both trained and untrained California sea lions. Footage was captured at the Smithsonian Zoological Park (Washington DC) using a stationary high-definition camera (GC-PX100BU, JVC, Japan) recording at 60 frames per second [ 13 ]. The large underwater viewing window allowed for observation over a distance of at least three body lengths as the sea lions swam passively. Additional recordings of trained sea lions performing directed maneuvers were collected at SeaWorld Florida and SLEWTHS at Moss Landing Marine Laboratories [ 33 ]. From these observations, four features were identified: a manipulable head (cervical), an articulatable pelvis (lumbar), hind flippers capable of acting as control surfaces, and fore flippers that serve both as primary propulsion and control surfaces (Fig. 5 A). To inform the design of robotic platforms, the swimming and maneuvering kinematics of California sea lions were analyzed, specifically focusing on the movements of their flippers and various body segments. The analysis was done using the two-dimensional digital tracking software Kinovea (Kinovea.org, France), which was used to measure rotations about different axes for the fore flippers, hind flippers, pelvis, and head. Video footage was selected where sea lions swam either parallel or perpendicular to the viewing window, ensuring clear visualization of specific movements. Kinematics were identified by analyzing videos that provided an unobstructed view of the targeted body regions and their respective angles (Fig. 6 ). Since each motion was captured from a single perspective, multiple videos from different angles were used to reconstruct a complete movement profile. The data was then aggregated and averaged to establish a standard motion, providing insight into the degrees of freedom required for each body segment. These findings directly informed the design of the actuators for the robotic platform. To model the 3D kinematics of fore flipper motion, an analysis was done that focused on the rotation at the base of the flipper during the characteristic propulsive stroke (Fig. 6 ). The motion was categorized into the three phases, involving roll f , yaw f , and pitch f rotations within the flipper’s frame of reference (Fig. 7 ). Throughout the stroke, the flippers moved through 140° of roll f , 110° of yaw f , and 110° of pitch f . The rotations during these motions were plotted, and key inflection points were identified as control points. To validate the stroke analysis, a Piecewise Cubic Hermite Interpolating Polynomial (PCHIP) spline was fitted to various control points, providing a smooth trajectory while ensuring that the fitted motion remained bounded in magnitude. This method allowed for twice-differentiable interpolation, facilitating realistic motion modeling [ 22 ]. The fore flipper kinematics were validated using the robotic system to match the motions observed by the animal. This decomposition was not intended to challenge existing descriptions of the fore flipper stroke, but rather to supplement prior findings with a formulation that could be directly implemented on the robotic platform for the present experiments. The bending of the fore flippers was identified as an important factor for propulsion and was therefore analyzed to ensure it could be accurately modeled in the robotic system. Based on the work of Friedman and Leftwich [ 13 ], along with additional tracking conducted for this study, it was found that most of the bending occurs at the wrist joint, with continuous flexion extending through the digits (Fig. 5 C), (Fig. 12 ). The amount of flipper deformation will influence both the direction and magnitude of the propulsive forces they generate. As a result, accurately modeling fore flipper bending will be critical to the development of the bio-robotic system. The same two-dimensional tracking technique was used to capture the motions of the head, pelvis, and hind flippers during a banking turn. Head motion was tracked from the body to the tip of the nose, while pelvic motion was measured from the body to the base of the tail. During the maneuver, the head was observed to roll a maximum of approximately 90°, yaw 40°, and pitch 90°, while the pelvis rolled 20°, yawed 30°, and pitched 80°. Throughout the turn, the hind flippers yawed 35° and rolled 60° [ 33 ]. It remained unclear whether the pitching of the hind flippers resulted from joint articulation or passive flexibility of the flippers. Together, these measurements offered important guidance for modeling the sea lion’s turning mechanics in a bio-robotic system. 4 Bio-Robotic System Design 4.1 Overview Based on the observations and analyses discussed in Section 2 and 3, the Stroke Experimentation and Maneuver Optimizing Underwater Robot (SEAMOUR) was developed as a versatile research platform. Its design enables investigations into various swimming and maneuvering strategies through the actuation of multiple control surfaces, including the head, fore flippers, pelvis, and hind flippers (Fig. 8 ). SEAMOUR was designed with features to proportionally match those of a sea lion, with a total length of 1.3m. The system features 14 actively controlled degrees of freedom (DOF), powered by high-torque waterproof servo motors (WR-7701, Xpert RC, USA). These include ± 60° of pitch and yaw at both the head and pelvis joints, 160° of roll f , 110° of yaw f , and 130° of pitch f at the fore flippers, and ± 30° of yaw with 160° of roll at the hind flippers. This arrangement was selected to enable precise, repeatable, and independently controllable motions, providing a highly adjustable platform capable of delivering consistent performance across a broad range of swimming and maneuvering experiments. SEAMOUR was developed as a modular research platform capable of producing biologically relevant motions while supporting a wide range of experimental configurations. To model sea lion body bending the head joint provides independent pitch and yaw actuation for modeling cervical bending, while the pelvic section uses the same mechanism to simulate lumbar motion. Each degree of freedom features independently adjustable actuation speeds and customizable motion profiles, allowing precise control of joint kinematics. The system’s modular design accommodates interchangeable fore flippers with adjustable shape and flexibility, along with interchangeable head shapes and tunable inertial properties. This adaptability allows SEAMOUR to evolve with new hypotheses and experimental needs without requiring a complete system redesign. The platform was engineered for reliable operation in both constrained and unconstrained testing environments, providing repeatable, consistent motions for controlled experimentation. Additionally, onboard sensing records the system rotations, enabling analysis of how the robot’s kinematics influence its body’s movement through the water. Certain design simplifications were made to prioritize the features being investigated in these initial studies. Continuous body bending in the thoracic section, biologically accurate head shape, and exact flipper scaling and contour were excluded from this iteration of the design to focus on isolating the effects of the included features. Although these other features likely influence swimming performance, their omission allowed for a more controlled evaluation of the targeted kinematic variables. Future work will expand both numerical and experimental platforms to examine the influence of additional joints, body flexibility, and alternative control surface geometries on swimming and maneuvering performance. 4.2 Main Body The main body of the robotic platform, representing the thoracic region of the animal, was designed to enclose all mechanical and computational components within a streamlined ellipsoid shape, closely resembling that of a sea lion [ 11 ]. Unlike the animal, this section was made rigid in the current iteration of the system and does not allow for continuous body bending. The outer shell of the body was composed of multiple vacuum-formed sections made from 1.5 mm high-impact polystyrene (HIPS), which magnetically attach to internal support ribs for easy access to the underlying hardware. The front and rear shell sections taper smoothly from the circular head and pelvic regions to the oval-shaped central section. An aluminum U-channel serves as the structural core, providing rigidity and mounting points for internal components. SEAMOUR’s main body is divided into two sections: the anterior section, which provides mounting for the head and fore flippers, and the posterior section, which contains a waterproof box housing the electronics and provides a mounting point for the pelvic section. A Raspberry Pi 4B (RPI) provided computational power and broad hardware compatibility, supporting motor control, sensor integration, and remote operation. A PWM-based daughter board (PN 2327, Adafruit, USA) was used for precise servo motor control, while a 9-DOF inertial measurement unit (IMU) (BNO085, PN 4754, Adafruit, USA) captured rotational and acceleration rates (Fig. 9 ). To enable remote operation, the RPI’s Wi-Fi antenna was modified so that it could be extended via a cable that breached the water’s surface. This antenna could either be dragged behind the robot using a surface float for continuous communication or remain attached to the outer shell after motor trajectories were loaded near the surface. The system was also designed with flexibility for future sensor integration, such as a depth sensor for active buoyancy control, a doppler velocity log for precise velocity and position measurements, or pressure sensors for local flow detection. Power for the system was supplied by a battery and distributed through two voltage regulators: one dedicated to the RPI and another for the motors. A 2250 mAh 3-cell LiPo battery (MaxAmps, USA) provided a minimum of two hours of continuous operation. The battery was potted in epoxy resin and fitted with waterproof connectors on both the power and balance leads, allowing it to be slid into the anterior section of the U-channel core. Power from the battery supplied a 5V 3.2A regulator (D36V28F5, Pololu, USA) for the RPI and a 15A regulator (D24V150F9, Pololu, USA) for the 14 servos (Fig. 9 ). The servo power regulator was modified to allow it to be powered on and off via commands from the RPI, while the motor regulator was modified to operate at 8.4V, enabling each motor to reach its maximum performance of 8.2 Nm of torque and rotational speeds of 428 degrees per second. As the interior of the robot is flooded to enable the modular design, all electrical components were housed within a waterproof box located in the posterior section of the robot. This box was fabricated using a FDM 3D printer and coated with multiple layers of polyurethane to ensure waterproofing. The top opening was sealed with a 6 mm thick laser-cut acrylic sheet and a continuous 6 mm square neoprene gasket, fitted into a shallow groove to create a knife-edge seal. The lid was fastened with 4 mm bolts that connected to heat-set threaded inserts embedded from opposite side to prevent pullout. Antenna and servo motor wires were epoxy-potted into the box’s sides, securely sealing all penetrations (Fig. 9 ). This design allowed SEAMOUR to operate at depths up to 4m. The system was designed so tests could be run while swimming freely or while attached to a float. Unconstrained tests were conducted utilizing an underwater docking station. This dock positioned the robot at a depth of 0.8 meters and featured two curved supports that matched the contour of the main body’s shell. Magnets embedded in a release rod connected to a steel bar running along the length of the posterior shell which allowed the system to be released remotely from the surface. This approach was used during both swimming and maneuvering tests conducted in this paper, which ensured a consistent starting position and orientation for each trial (Fig. 10 ). For constrained testing, the system was designed to be attached to a surface float, restricting its motion to the xy plane while preserving x axis rotations. To accomplish this a Ø160mm FDM 3D-printed Delrin roller bearing was fabricated and positioned between the front taper and main body shells. By attaching the system to a float, its position and orientation can be tracked at the surface and a surface docking system can be implemented for repeatable trials. This approach was not used in the studies presented in this paper, but its utility will be important for future studies. 4.3 Fore Flippers The fore flippers were designed to model the fundamental motions and bending of the biological sea lion flippers during characteristic swimming and turning maneuvers. Their motion was modeled at the elbow joint of the animal and actuated using three servo motors per flipper, providing independent control of pitch f , yaw f , and roll f . Described in the fore flipper frame of reference, the pitch f motor attaches to the center support structure, next in the chain is the yaw f motor, and lastly the roll f motor which acts as the final connection point for the fore flipper (Fig. 11 ). To increase the rigidity of the assembly, bearings were added on axis to the pitch f and roll f motors to support the radial loads. This assembly accommodates flippers of varying sizes and degrees of flexibility, while the available DOF enables a wide range of motions, ensuring the system is not limited to strictly biologically derived movements. For example, rigid flippers were used in place of the standard flexible flippers during the experimental trials conducted to validate the numerical model of the system, as discussed in Section 4.8 [ 19 ]. SEAMOUR’s fore flippers were designed to model the general shape, proportions, and placement on the body. Certain biological characteristics, such as the trailing edge crenelations [ 23 ] and the specific hydrofoil shape [ 12 ] were simplified in the robotic system as these features were not the focus of these studies. The fore flippers were positioned to reflect the natural anatomy of a sea lion, but their size was reduced by 17% relative to a proportional scale to accommodate motor torque and speed limitations while maintaining key functional trends associated with propulsion and maneuvering. Preliminary tests showed that the hydrodynamic lift generated by the modeled fore flippers where similar to the actual sea lion fore flipper over a range of angles of attacks (unpublished data). The bending of the fore flippers was modeled based on the bending profile observed during the characteristic stroke. Each flipper was designed with a rigid base and a gridded support structure that enabled both span-wise and chord-wise bending to be modified during the design process (Fig. 13 ). The proportions of the rigid and flexible sections were derived from the animal's anatomy, where the forearm represented two-fifths and the hand portion represented three-fifths of the flipper (Fig. 5 B) [ 12 ]. This design ensured the fabricated flippers modeled natural bending from the wrist joint through the finger-like structure [ 10 ], [ 13 ]. Bending of the fore flipper was passively induced by self-loading during movement through water, with the degree of bending adjustable by modifying the taper of the grid pattern or the number of grid sections. Finite element analysis (FEA), conducted using SolidWorks CAD software (SolidWorks, USA), simulated the response of the flipper's support structure to a non-uniform distributed 5N load, modeling the flipper's motion through water (Fig. 12 ). This simulation facilitated fine-tuning of the flipper’s structure before fabrication and validation on the mechanical system. The analysis predicted a tip displacement of 0.103 m, which when scaled closely matched the bending characteristics observed in the biological system (Fig. 12 ). The fore flippers were fabricated using a multi-step casting process (Fig. 13 ). First, a positive mold was created using an FDM 3D printer (F120, Stratasys, Minnesota USA). This mold was then vacuum-formed (450DT, Formech, USA) with clear 1.5 mm PETG sheets to produce the two halves of a negative mold. The 3D-printed support structures were positioned using a square mounting rod, which later served as the connection point to the motor assembly. The clear molds were fitted with a neoprene rubber gasket and secured using a 3D-printed framing clamp and bolts. A two-part silicone with a shore hardness of 30A (Dragon Skin 30, Smooth-On, USA) was poured into a preformed funnel at the top of the mold. The transparent molds allowed for visual confirmation that no air pockets were present in the silicone and that the support structure was properly aligned. 4.4 Pelvis The pelvic section was designed to emulate the bending of the animal’s lumbar region using a two-axis gimbal mechanism driven by two motors. This setup provides independent control over pitch and yaw movements. The gimbal is enclosed within a half-dome structure, with its axes aligned to a shared center of curvature, allowing up to ± 60° of articulation from the midline in each direction. As minimal pelvic roll rotation was observed by the animal, rotation about this axis was excluded from the joint design. The pelvic section also integrates the motors responsible for hind flipper actuation and voids throughout are used for ballast material. 4.5 Head SEAMOUR’s head was mounted on a two-axis gimbal system, modeling the degrees of freedom of a sea lion’s cervical region. In these initial studies, the head was axisymmetric to reduce mechanical complexity, measuring 0.265 m in length with a cross-sectional area of 0.02 m². To minimize drag and mitigate potential flow disruptions caused by head articulation, a four-way stretch spandex fabric was secured between the head and the tapered body shell (Fig. 14 A). The head mounting system was designed for interchangeable head shapes and sizes, allowing for future investigations into the effects of head asymmetry. For example, a smaller axisymmetric head is utilized in Section 6.6, demonstrating this modularity. Beyond its adaptable form, the head also accommodates ballast material and space for a future vision system. 4.6 Hind Flippers The hind flippers of the bio-robotic system were designed to model the kinematics of the animal using flexible flippers that could yaw and roll independently. Housed within the pelvic section, the hind flipper motors enable each flipper to yaw a maximum of ± 30° from midline and roll 160° (Fig. 14 B). The flexible flippers were fabricated in the same manner as the fore flippers; however, their bending characteristics were not explicitly characterized, as the influence of their flexibility was not explored during these studies. When comparing the modeled hind flippers to the actual animal the hydrodynamic lift generated over a range of angles of attacks were similar (unpublished data). 4.7 Trim Lead weights and extruded polystyrene foam were added to designated voids in the head, body, and pelvis to adjust the locations of the center of gravity (COG) and center of buoyancy (COB), achieving neutral buoyancy and a horizontal equilibrium in the water. These adjustments positioned the COB above the COG in both the x-z and y-z planes (Fig. 15 A), creating moments that affected the system's roll and pitch stability [ 36 ]. The head, body, and pelvis sections were individually trimmed for neutrally buoyancy before being assembled. Trim modifications were also used to shift the COG and COB of the overall system anteriorly or posteriorly, with efforts made to align their positions close to the fore flippers as seen in the biological model [ 18 ]. Each one of these adjustments were updated in the 3D CAD model of SEAMOUR (Creo Parametric 10.0, USA), which was then used to calculate the COG of the system. In the final configuration, the COG for the entire system was calculated to be -0.07 m in the x-direction from the coordinate frame (Fig. 15 B), which places it 0.12 m in front of the system's overall center. To calculate the COB, the solid parts and any air pockets in the CAD model needed to be set to a uniform density. Making these adjustments would allow the COG calculator in the 3D CAD software to be used to find the centroid of the displaced volume of fluid (i.e. COB). This calculation showed that the COB was also located − 0.07 m in the x-direction from the coordinate frame. 4.8 Numerical Model Complementary numerical models of SEAMOUR were developed to simulate, analyze and visualize the coordinated flipper and body motions of the robotic system in water [ 19 ], [ 37 ]. Utilizing both Euler-Poincaré and Newton-Euler formulations, these models are used to investigate and evaluate the bio-robotic system's maneuverability and stability across different swimming modes. To simulate the fluid forces, drag, lateral, lift and added mass forces were incorporated into the numerical model. The hydrodynamic coefficients, essential for calculating these forces, were determined through computational fluid dynamics (CFD) simulations and analytical techniques such as strip theory. These models were validated against SEAMOUR for various swimming and maneuvering trials. These models allow for the examination of the system's maneuverability, exploration of new strategies for propulsion, and assessment of design modifications prior to implementation on SEAMOUR. 5 Experimental Procedure and Data Collection 5.1 Overview Experiments were conducted to implement and evaluate various swimming and maneuvering strategies using the bio-robotic system. These tests were designed to characterize the system’s dynamics, assess the use of various control surfaces, and identify performance trends as the robot moved freely through the water. Prior to testing, the system was trimmed to be horizontally level, with individual body sections trimmed for neutral buoyancy. The first set of experiments examined how the systems current COG and COB influenced its ability to return to equilibrium after being displaced at various angles (N = 120), with stability assessed by measuring the response and settling times as the body realigned. The second set of experiments evaluated how different combinations of fore flipper and body motions could be used to achieve rectilinear swimming (N = 15), with body orientation and x- and z-axis translations recorded to assess each stroke’s ability to maintain straight and level trajectories. The third set of trials investigated the use of head and pelvis motions to perform controlled pitch and yaw maneuvers (N = 60), which are assessed by recording body orientation to quantifying angular positions and rates as control surfaces were actuated through a range of angles. Finally, insights from these experiments were combined to develop coordinated swimming and maneuvering sequences, demonstrating the full versatility and modularity of the system. 5.2 Passive Roll and Pitch Stability The impact of the system's COG and COB locations on its roll and pitch stability were determined by observing the body's angular rate as it returned to equilibrium. Experiments were conducted by manually positioning the robot at various roll and pitch angles before releasing it. An onboard IMU recorded the resulting angular changes over time at a sampling rate of 20 Hz (Fig. 16 ). Tests were conducted with roll and pitch angles of 20°, 40°, 60°, and 80°. Pitch tests were performed with the system in a streamlined position, with the fore flippers held against the body. Roll tests were conducted under two conditions: with the fore flippers streamlined and with the fore flippers extended ( \(\:{\mathbf{y}\mathbf{a}\mathbf{w}\mathbf{e}\mathbf{d}}_{\varvec{f}}\) and \(\:{\mathbf{r}\mathbf{o}\mathbf{l}\mathbf{l}\mathbf{e}\mathbf{d}}_{\varvec{f}}\:\) 90°). The robot was tested by rolling it to the left and right, as well as pitching it in head-up and pelvic-up orientations. The data was offset to align t = 0 with the release time and trimmed to 16 seconds. The systems pitch or roll orientation from three trials were averaged for analysis and the standard error between trials was calculated. Settling time was identified as the first instance when the signal entered and remained within a defined threshold band around the final value. This final value was estimated by fitting a linear trend line to the last 7 seconds of the roll data and extrapolating to the endpoint. The threshold band was set as ± 0.1° from the final value, allowing for small fluctuations while capturing the signal's convergence behavior. The algorithm checked for continuous periods within this band that lasted for a minimum duration of 0.10 seconds to confirm stability. The first point at which the data satisfied these criteria was recorded as the settling time. Due to the differences in the data set, a different methodology for calculating the settling time for pitch tests was needed. First the average of the final 10 data points of each test was calculated to represent the steady-state value. A ± 0.1° band was established around this steady-state value to define the acceptable range for settling. The settling time was identified as the first time point where the data enters this band and remains within it for the rest of the duration. This approach ensures that the system is considered settled only when it consistently stays close to its final value. These two approaches for determining settling time were necessary because SEAMOUR could not be precisely trimmed to achieve equilibrium at 0° in both the x and y axes. 5.3 Swimming To evaluate the impact of flipper and body motions on SEAMOUR's swimming capabilities, studies were conducted to assess various approaches for achieving rectilinear swimming. Three swimming techniques were tested for their effectiveness in producing straight and level locomotion: a biologically derived characteristic stroke (implemented both with and without pelvic pitching) and a modified side paddle stroke (Fig. 18 ). Developing a stroke that minimizes pitching and vertical displacement while maximizing forward velocity is essential for maintaining a steady trajectory and executing maneuvers at speed. The kinematics for the characteristic stroke were derived from existing literature, in-house kinematic investigations, and preliminary studies. While a separate future publication will detail a study investigating the lift and thrust forces produced during different phases of the characteristic sea lion stroke, the findings from that work provided kinematics that minimized lift and produced optimal thrust for these experiments. Additional trials utilized the pelvic section as a control surface to correct for pitching during the characteristic stroke. Preliminary tests determined that pitching the pelvic section downward by 15° during the stroke transition (between the power and paddle phases) and gradually returning it to a streamlined position during the recovery phase produced effective results. The modified side paddle stroke was designed to utilize only the fore flippers for rectilinear swimming. By manually adjusting the PCHIP spline control points for fore flipper motion, a new swimming gait was developed. This gait was adapted from the best-performing reinforcement-learned gait evaluated in earlier work [ 37 ], with slight tuning to accommodate design modifications. This modified stroke replaces the power phase with a side-body paddle motion, allowing control of vertical body movement through flipper roll and pitch adjustments. During each experiment, the position of SEAMOUR was tracked using a stationary underwater camera while orientation was recorded using an on board IMU. Each stroke type had a three second period with each trail consisting of four consecutive strokes. A GoPro Hero 11 camera (GoPro, USA), positioned to capture the side of the robot as it moved through the water, was set 8 m away and to record at 24 frames per second in 4K resolution using a wide frame of view. The video footage was then post-processed to correct lens distortion, lighting, and color (Adobe Premiere Pro, USA) before being tracked in Kinovea. To calibrate the scale of the video frame, a known length on SEAMOUR was set using the software. Pitch orientation data was collected during each trial using the IMU, which recorded at 20 Hz starting two seconds before and continuing throughout the trial; with the initial reading used to correct for any offset in the robot’s starting orientation. The position and orientation data from each study were aligned and results averaged across three trials, with the variability between them recorded as standard error. 5.4 Yaw and Pitch Turns To evaluate the effectiveness of pitch and yaw turns on SEAMOUR, various combinations of the head and pelvis angles were explored. During these tests, SEAMOUR swam unconstrained, using its fore flippers to reach an initial velocity of 0.31 m/s at the start of each trial. Following this, the head, pelvis, or both were actuated to either 30° or 60° in yaw or pitch. During the combined tests, the head and pelvic were actuated symmetrically, i.e. head at 30° pitch and pelvic at 30° pitch. Pitch and yaw tests were performed about their respective axes—pitch about the y-axis and yaw about the z-axis. Each pose was held for 5 seconds, and the IMU recorded the system’s orientation. A representative data set of three trails was then averaged, with the standard error between trials calculated. During each pose, the fore flippers remained in a streamlined position alongside the body, and the hind flippers were held at a roll angle of 0°. 6 Results and Discussion 6.1 Overview SEAMOUR was successfully developed as a research platform to investigate the use of various biological and non-biological swimming and maneuvering gaits. Across all experimental trials, the system demonstrated consistent, repeatable performance with low positional and angular standard errors, validating SEAMOUR’s reliability and versatility as a modular, multi-bodied robotic platform. Initial static stability tests established a baseline, revealing that roll response varied with fore flipper configuration, as extended flippers improved stability by reducing angular velocity, settling time, and overshoot, while pitch maneuvers exhibited lower angular velocities but comparable settling times to roll tests with streamlined flippers. When implementing a biologically derived swimming stroke, known as the characteristic stroke, the system saw significant translation in heave and high pitch changes through the trials. When the pelvis was incorporated during the characteristic stroke, heave translation was mitigated, due to reduced pitching, and surge translation was greatly improved. Further refinement using a modified side paddle stroke eliminated the need for pelvic actuation while still achieving effective forward translation and maintaining a low body pitch angle. In maneuvering trials, the pelvic section had the greatest influence during pitch and yaw maneuvers, while combining head and pelvic articulation saw slight improvements. Overall, pitch-based maneuvers consistently outperformed yaw-based maneuvers under all test conditions. These findings collectively highlight the ability to implement various swimming and maneuvering gaits, which would pave the way for additional gait strategies tailored to specific locomotory objectives. 6.2 Influence of COG and COB As SEAMOUR is a multi-bodied robot, the buoyancy characteristics of its head and pelvic sections can influence the performance of the overall system when articulated. If these independent sections were negatively buoyant, while the overall system was neutrally buoyant, any displacement of these sections from the body’s midline would have caused the COG to shift more than the COB, causing the system to roll. To mitigate this action, the head and pelvis sections were individually made neutrally buoyant to minimize the influence of hydrostatic forces, particularly during the maneuverability studies. This configuration ensures that the COG and COB approximately shift in unison when individual body sections are articulated. The COG, COB, and moments of inertia are closely interconnected, and their combined configuration significantly affects the maneuverability of the bio-robotic platform. In a configuration where the head and pelvic sections are negatively buoyant while the overall system remains neutrally buoyant, the moments of inertia about the y and z axes will be greater compared to a configuration where each section was independently neutrally buoyant. The former configuration resembled a dumbbell, while the latter was akin to a pipe. Consequently, a negatively buoyant head and pelvis increase the system’s pitch and yaw stability, making it suitable for applications prioritizing stable straight swimming. However, this setup is less optimal for a maneuverable system, further justifying the trimming of the COG/COB to enhance maneuverability. Understanding and optimizing the moments of inertia are essential in the design of effective underwater vehicles, as the offset and relative positions of the COG and COB directly influence a system’s stability and the ability to roll, yaw, and pitch. These parameters can be strategically adjusted to achieve specific performance characteristics depending on the operational goals of the vehicle. Future work will investigate the effects of modifying these properties through additional testing. 6.3 Passive Roll and Pitch Stability During tests when the system was rolled away from equilibrium, its response varied significantly based on the fore flipper configuration but demonstrated consistent trends across different initial roll angles (Fig. 18 ). Larger initial roll angles resulted in greater overshoot magnitudes when the fore flippers were streamlined, whereas the opposite occurred when the fore flippers were extended. By extending the fore flippers, the angular velocity during the first second is decreased by 66% across all four angles (Table 1 ). The mean standard error across both roll configurations ranged from ± 0.17° to ± 0.47° across each 16 second trail (Table 2 ), demonstrating consistent performance. With the fore flippers in the streamlined configuration, the roll angle underwent multiple cycles of damped oscillations before settling, while in the extended configuration, stability was achieved after a single damped oscillation. The settling time in Fig. 18 A and B is denoted by a black dot on each line. This variation in the system's response can be attributed to the difference in drag between configurations. Notably, when comparing the two roll configurations, the settling time was significantly reduced when the fore flippers were extended (abducted). Comparing the streamlined and extended positions revealed that extending the fore flippers reduced settling time by 53% for 40°, 53% for 60°, and 49% for 80° initial roll angles. With the fore flippers extended during the roll test the settling time for release angles of 40°, 60°, and 80° consistently fell between six and seven seconds, while the 20° release angle deviated from this pattern. This was likely due to the system's low restoring moment at smaller initial angles, causing it to settle slightly off the target position after the first overshoot. Consequently, the extrapolated trend line predicted a final value that was farther from the expected position at 15.3s. Table 1 Mean angular velocity during the first second of each trial for fore flipper streamlined and extended configurations at 20°, 40°, 60°, and 80° roll angles from equilibrium. Mean angular velocity Fore Flipper Configuration 20° (°/s) 40° (°/s) 60° (°/s) 80° (°/s) Streamline 30 49 61 70 Extended 10 16 22 24 Table 2 Mean standard error for fore flipper streamlined and extended configurations at 20°, 40°, 60°, and 80° roll from equilibrium. Mean standard error Fore Flipper Configuration 20° ( ± °) 40° ( ± °) 60° ( ± °) 80° ( ± °) Streamline 0.18 0.17 0.22 0.47 Extended 0.33 0.42 0.24 0.35 When the robot was pitched away from its equilibrium and released, it had lower mean angular velocities compared to both roll tests (Table 3 ), but had comparable settling times to the roll configuration with the flippers in the streamlined position (Fig. 19 ). The increase in drag and moment of inertia during pitch tests resulted in lower angular velocities, preventing the system from overshooting its final value in this orientation. Over the 16 second trials the mean standard error ranged from ± 0.66° to ± 1.26° (Table 4 ). Table 3 Mean angular velocity during the first second of each trial for fore flipper streamlined and extended configurations at 20°, 40°, 60°, and 80° pitch angles from equilibrium. Mean angular velocity Fore Flipper Configuration 20° (°/s) 40° (°/s) 60° (°/s) 80° (°/s) Streamline 0.86 1.89 2.68 4.84 Table 4 Mean standard error for fore flipper streamlined and extended configurations at 20°, 40°, 60°, and 80° pitch from equilibrium. Mean standard error Fore Flipper Configuration 20° ( ± °) 40° ( ± °) 60° ( ± °) 80° ( ± °) Streamline 1.26 0.66 0.82 1.20 The results demonstrated that SEAMOUR’s roll response resembled an underdamped system, while its pitch acted more like a critically damped system. This response was consistent with the smaller moment of inertia about the roll axis and the reduced drag encountered in that direction. As roll maneuvers will not be explored in these studies, a higher static roll stability will help isolate the effects of control surfaces on pitch and yaw maneuvers. Additionally, the low static pitch stability ensured that pitch maneuvers remain largely unaffected, allowing for a clearer evaluation of rectilinear swimming performance in terms of both thrust and lift without significant counteracting effects from static pitch stability. The evaluation of SEAMOUR's passive roll and pitch stability established a critical baseline for understanding its inherent static stability, which must be assessed before investigating the system’s maneuvering abilities, particularly when modifications are made to its COG and COB. This baseline served as a reference for assessing how static stability influenced the system’s performance during dynamic tests. Specifically, these investigations help determine whether stability enhances or constrains the effectiveness of the control surfaces. This distinction is key to the subsequent discussion on the system’s maneuvering capabilities and the role of static stability in shaping overall performance. 6.4 Swimming When a biologically derived characteristic stroke was implemented on SEAMOUR, pitching and heave translation affected its other all surge translation. Over the course of four strokes, the system pitched up by a total of 36°, with an average of 23°. This stroke also saw heave translation of 0.38m (± 0.03 m) and surge translation of 2.89 m (± 0.08m) (Fig. 20 ). The observed pitching resulted from the lift forces generated in combination with the COG location, due to a moment arm. This shows that pitch stability can be particularly critical during propulsive strokes. Since SEAMOUR’s static pitch stability is low, this stroke easily overpowered it. As the system continued pitching upward, increased surface area was exposed to the forward flow, further exacerbating pitching in subsequent strokes. The performance of the biologically derived stroke on SEAMOUR highlighted both the stroke’s limitations in straight swimming and its potential for maneuverability. Due to key differences between SEAMOUR and a biological animal, the stroke did not achieve effective straight swimming. However, it demonstrated how fore flipper lift forces generate pitching moments, which may be useful for maneuvers. Fortunately, the fore flippers are not restricted to a single set of kinematics, allowing their trajectories to be modified or additional control surfaces to be incorporated during the stroke to achieve straight swimming (Fig. 21 ). To minimize pitching and heave translation during the characteristic stroke, the pelvic section was used during the stroke, resulting in better surge translation and less overall pitching. This stroke maintained an average pitch angle of 9° and reduced the final pitch angle by 63% after four strokes, compared to trials without pelvic actuation. The system achieved a final horizontal translation of 3.24m (± 0.09 m) and a vertical translation of -0.05m (± 0.04m). These results highlight how the coordinated use of additional control surfaces during a fore flipper propulsive stroke can balance forces and produce straight swimming. The modified side paddle stroke had the least amount of pitching and comparable surge translation as the characteristics stroke with the pelvic actuation. This stroke saw horizontal translation of 3.23 m (± 0.08 m) while maintaining an average pitch angle of 2° over four strokes. An increase in pitch angle between seven and twelve seconds was observed, potentially due to drag from the surface antenna. Slight negative buoyancy was evident, as indicated by minimal pitching during the first three strokes while maintaining a consistent downward trend in the Z direction, resulting in a final vertical translation of -0.16 m (± 0.06 m). This approach demonstrates how modifying the 3D kinematics of the fore flipper can achieve effective straight swimming, and by eliminating the need for additional control surfaces during fore flipper propulsion, these surfaces can be reallocated for coordinated maneuvers, as discussed in Section 6.6. Low standard errors were observed in position and orientation data during various swimming and maneuvering trials. As the system swam unconstrained for 12 seconds, covering 3 meters of horizontal translation, the standard error for both horizontal and vertical displacement remained below 8% of SEAMOUR’s total length. During these trials the standard error for the pitch angle was 1.5° for the characteristic stroke, 0.8° for the characteristic stroke with pelvic actuation, and 2.2° for the modified paddle stroke. These low error margins demonstrate the system’s ability to have repeatable swimming performance across different gaits and configurations. 6.5 Pitch and Yaw Maneuvers This section evaluates the system's trends and performance impacts. While a comprehensive quantitative analysis is reserved for a future publication, the results presented here highlight key trends in the system's performance. Quantitative data, specifically standard errors, are used to demonstrate the system's repeatability and consistency across different maneuvers and test conditions. During pitch maneuvers, simultaneously actuating the head and pelvis at higher angles yielded greater mean rotational rates and final angular positions, with the pelvis contributing more significantly than the head (Table 5 ). Comparing final angular positions, the pelvic section consistently outperformed the head section at both tested pose angles. Increasing the actuation angle from 30° to 60° produced a similar increase in rotation for the pelvis and combined sections, while the head section exhibited a more modest improvement. When the head and pelvic sections were actuated together, the resulting angular positions closely matched those of the pelvic section alone, with only minor additional increases at both pose angles. For yaw maneuvers, the pelvic section consistently outperformed the head section when actuated independently, while the combination of both produced the highest final angular positions. The pelvic section achieved substantially greater rotations than the head at both tested pose angles, and when actuated together, the head and pelvic sections further improved performance beyond the pelvic section alone, with more noticeable gains at the lower pose angle. Increasing the actuation angle from 30° to 60° produced varied outcomes: the head section showed minimal improvement, while both the pelvic and combined sections demonstrated substantial increases in final angular positions. In these yaw studies, the head contributed more effectively when paired with the pelvic section. This could be attributed to the hind flippers' position during the maneuver. Since they were rolled flat in these tests, they cut through the water, reducing the pelvic section's effectiveness and allowing the head's motion to exert a relatively greater influence on the generated forces. When comparing pitch and yaw maneuvers, pitch consistently outperformed yaw in both final angular position and average turning rates (Fig. 22 ) (Table 5 ). Head articulation at both pose angles resulted in relatively low final angular positions for both pitch and yaw. In contrast, pelvic articulation during pitch produced noticeably higher final positions than yaw at both tested angles. When the head and pelvis were actuated together, pitch maneuvers demonstrated greater increases in final angular position compared to yaw, with larger gains observed at the higher pose angle. Average turning rates for head articulation were similar between pitch and yaw. However, pelvic articulation favored pitch, with notably higher turning rates at both pose angles. When combining the head and pelvis, pitch maneuvers again produced substantially higher turning rates compared to yaw, with the difference more pronounced at the lower pose angle. Detailed values for mean turning rates are presented in Table 5 . Clear trends were evident across both pitch and yaw studies, depending on which sections were actuated. During all maneuvers, the pelvic section consistently outperformed the head, likely due to the axisymmetric shape of the head and the inclusion of large, flat hind flippers on the pelvic section, which additionally enhance its effectiveness during pitch maneuvers. For both 30° and 60° maneuvers, the pelvic section alone initially outperformed the combined actuation of the head and pelvis. In the first 0.5 seconds, the pelvic section achieved a larger initial angular change compared to the more gradual response of the combined pose. This trend was observed in both pitch and yaw studies, with a more pronounced effect during yaw tests. This pattern becomes clearer when considering the results of head actuation alone, which consistently produced an initial negative angular position in both pitch and yaw maneuvers before transitioning to a positive angle. During head actuation, the motor, located anterior to the system's COG, imparts a reaction moment. Consequently, the main body, containing the IMU, experiences an initial negative angular displacement. This effect can be seen most dramatically in the head pitch and yaw 60° trials. In contrast, pelvic section actuation induces the opposite effect, which can be seen most prominently in the pelvic 60° yaw plot. Unlike the yaw studies, which show a relatively linear increase, the pitch studies are influenced by restorative forces resulting from the offset between the COG and COB. The findings revealed low standard errors across trials and configurations, demonstrating consistent trends. Since standard error propagated over the length of the trials, examining the error at the final 3 seconds provided insight into overall performance. At 3 seconds, the final pitch error was ± 2.0° and ± 0.9° for head-alone trials at 30° and 60°, respectively, ± 0.4° and ± 0.3° for pelvis-alone trials, and ± 0.3° and ± 0.4° when the head and pelvis were used together. For yaw tests, the final error for head-alone trials was ± 2.4° and ± 2.0°, for pelvis-alone trials was ± 2.5° and ± 3.1°, and for combined actuation was ± 3.7° and ± 0.8° at actuation angles of 30° and 60°, respectively. These results indicate that the system maintained a low error range, varying from as little as ± 0.3° to a maximum of ± 3.7° over the three-second turning period as it maneuvered through 3D space, demonstrating the system’s repeatable performance. Table 5 Mean turning rate for pitch and yaw trials over the 3s trails Mean Turning Rates Pose Pitch (°/s) Yaw (°/s) Head 30 3.17 4.12 Head 60 5.23 5.33 Pelvic 30 12.69 8.38 Pelvic 60 22.99 13.45 Head and Pelvic 30 11.88 5.51 Head and Pelvic 60 21.77 11.67 6.6 Coordinated Maneuvers Insights gained from the experimental swimming and maneuvering trials provided a foundational understanding to produce coordinated maneuvers using flippers and various body segments. These coordinated maneuvers used various forms of flipper and body motions to achieve roll, yaw, and pitch maneuvers. These maneuvers followed pre-prescribed motor kinematics, meaning the system operated without feedback during its movements. For each test, SEAMOUR was positioned at a depth suitable for the intended trajectory. SEAMOUR's versatile maneuvering strategy, involving coordinated movements of its head, pelvis, and fore flippers, allowed it to effectively follow a complex curvilinear path. Figure 24 illustrates a 20-second sequence with the robot’s position recorded every 4 seconds and overlayed onto a single image. To execute this sequence, SEAMOUR first accelerated using two modified side paddle strokes. The head and pelvis were then pitched down 30° while another stroke to initiate a dive. With the system pitched downward, the head and pelvis returned to a streamlined position, and an additional stroke was used to drive SEAMOUR deeper at a 25° angle. Subsequently, the head and pelvis were pitched up 30° during an additional stroke to change direction upward. Following this, the head and pelvis returned to the streamlined position, and another stroke was performed to continue the ascent at a 40° angle. To level the body, the head and pelvis were pitched down 30° during a subsequent stroke, after which SEAMOUR completed additional modified side paddle strokes to swim out of the camera’s view. This sequence demonstrates the effective use of the head, pelvis, and fore flippers to perform a coordinated series of maneuvers so swim through a curvilinear path. By leveraging the modularity of the system, adjustments to head size and the COG/COB distribution were made to reduce roll stability to execute the three coordinated maneuvers seen in Fig. 25 . The head and pelvic sections were designed to be negatively buoyant while maintaining overall neutral buoyancy. This configuration caused the COG to shift laterally when these sections yawed, generating restorative forces that rolled the body to vertically realign the COG and COB. The z distance between the COG and COB was reduced while still retaining the system horizontal orientation in water. Additionally, a smaller axisymmetric head was added as less volume was required for these hydrostatic adjustments. SEAMOUR demonstrated a 360° roll maneuver using only a single flipper, as illustrated in Fig. 25 A. To execute this, the robot first accelerated with three modified side paddle strokes. Immediately following, a single flipper executed three consecutive strokes. These strokes leveraged the fore flipper's roll angle during both its abduction and adduction phases, resulting in a rotation about the bodies roll axis. Beyond this, traditional asymmetric fore flipper roll angles have also been explored for body rolling, a technique used in the coordinated maneuvers shown in Fig. 25 B and Fig. 25 C. SEAMOUR executed a biologically derived banking turn (Fig. 25 B), modeling the maneuver observed in the California sea lion (Fig. 4 ). The maneuver began with the fore flippers \(\:{\text{y}\text{a}\text{w}\text{e}\text{d}}_{f}\) 90° and asymmetrically \(\:{\text{r}\text{o}\text{l}\text{l}\text{e}\text{d}}_{f}\) , creating an effective offset of 40°, where one flipper is angled with the top face into flow and the other with the bottom face in the direction of flow. Simultaneously the head and pelvic section were yawed 60° while the hind flippers were rolled flat and yawed outward by 20°. As the body rolls, the head and pelvic sections adjusted their yaw and pitch angle to maintain alignment with the xy plane, reaching 60° of head/ pelvic pitch when the body is rolled 90°. The fore flippers were then \(\:{\text{r}\text{o}\text{l}\text{l}\text{e}\text{d}}_{f}\) to 90° during the remainder of the turn. To return the body to its initial state, the head and pelvic sections were streamlined as the lower fore flipper was flapped to roll the body, and the other fore and hind flippers slowly return to their streamlined position. This highlights SEAMOUR’s ability to execute maneuvers modeled after the sea lion, allowing for deeper investigation into biological approaches. In Fig. 25 C, SEAMOUR executes a 180° pitch turn by performing a front flip type maneuver. This maneuver was achieved by first using the characteristic stroke with pelvic actuation to accelerate the system. Subsequently, the head and pelvic sections were pitched downward by 60° as five modified side paddle strokes were executed to flip the body 180°. With the fore flippers then extended, they were \(\:{\text{r}\text{o}\text{l}\text{l}\text{e}\text{d}}_{f}\) asymmetrically, causing the body to roll 180° back to equilibrium before performing further rectilinear swimming strokes. These coordinated maneuvers highlight SEAMOUR's versatility and demonstrate the successful integration of key findings from these studies. 7 Conclusion and Future Work A bio-robotic underwater system, inspired by the California sea lion, was successfully developed and validated as a capable and reliable research platform. Through a series of comprehensive swimming and maneuvering experiments, this paper has demonstrated the system's foundational design and its effectiveness as a research tool. These tests showcased the platform's versatility and repeatability, supported by a modular design built for expanding hypotheses. The system uniquely facilitates both independent and combined evaluations of various factors contributing to its performance, as demonstrated through coordinated swimming and complex roll, yaw, and pitch maneuvers. Beyond exploring biologically derived swimming gaits, the platform can also be tuned to investigate diverse hand-tuned or reinforcement-learned gaits. Ultimately, SEAMOUR stands as a versatile and reliable platform specifically designed to advance the exploration of morphological and kinematic traits contributing to effective underwater swimming and maneuvering. While the system achieved a maximum velocity of 0.31 m/s during the swimming trails, SEAMOUR’s modular design allows for modifications that can increase swimming speed. Several approaches can be explored, such as developing swimming gaits specifically tailored to the robotic platform, designing flippers with larger surface area, and replacing the current motors with models offering higher speed and torque capabilities. Additionally, incorporating supplementary control surfaces to assist with propulsion could further increase the system’s velocity. Future work will investigate modifications to optimize the platform’s swimming performance for a broader range of experimental applications. To gain a deeper understanding of the impact of various control surfaces on pitch and yaw maneuvers, a more detailed investigation is necessary. By constraining the system to so that yaw and pitch turns can be conducted on the same plane, eliminating any restoring moments during pitch tests, a better comparison between pitch and yaw turns can be facilitated. Expanding the range of actuation angles would also provide valuable insights into how these control surfaces influence maneuvering capabilities. Additional tests should also assess the contribution of the fore flippers on turning performance. To further evaluate the effectiveness of pitch and yaw maneuvers, position data should be collected, enabling the calculation of turning radii for each maneuver and providing a more comprehensive understanding of the system's maneuverability. Having a better understanding of the system’s maneuverability will provide valuable insights into the system's potential in dynamic flow environments. These evaluations can provide insights into how specific features of the animal's swimming and maneuvering can offer potential strategies to enhance the maneuverability of UUVs. Active buoyancy control should be integrated into the system's design to enable tests that isolate experimental factors and yield more repeatable results. During the current testing, constant buoyancy changes presented challenges. This was primarily because many components, fabricated using FDM 3D printing, experienced gradual water saturation, leading to a shift toward negative buoyancy over time. To compensate, slight adjustments were manually made to the main body's buoyancy using foam throughout the studies. Future enhancements should incorporate an active buoyancy management system, which would either maintain consistent buoyancy during operation or enable lateral buoyancy control for pitching the system. Declarations Acknowledgments This research was funded by the Office of Naval Research (Dr. Tom McKenna, Program Officer, ONR Code 341). The funder played no role in study design, data collection, analysis and interpretation of data, or the writing of this manuscript. We would like to thank the work conducted by students in the Liquid Life Lab at West Chester University, as well as the Biologically Inspired Energy Laboratory at George Washington University. The authors would like to thank Anthony Paul Bibeck and Ahmet Yalim Kiral for their contributions in conducting bio-robotic experiments and collecting valuable data. Data Availability The raw datasets generated and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request. Code availability The underlying code for this study is not publicly available but may be made available to qualified researchers on reasonable request from the corresponding author. Author Contributions N.M designed and developed the robotic system. N.M., S.K., and A.D. conducted studies with the robotic system. F.E.F provided biological insights to the California sea lion. M.C.L. provided information on the fluid dynamic interaction of the fore flippers. H.G.K worked with S.K on the development of the numerical models of the robotic system. J.L.T advised the building and testing of the robotic system. All authors read and approved the final manuscript. Competing Interests The authors declare no competing interests. Additional information Correspondence and requests for materials should be addressed to Nicholas Marcouiller ( [email protected] ). References “ONR Technology and Research,” Office of Naval Research. Accessed: Feb. 25, 2025. [Online]. Available: https://p74l1103a01.dc3n.navy.mil/our-research/onr-technology-and-research “New Blue Economy | National Oceanic and Atmospheric Administration.” Accessed: Feb. 25, 2025. [Online]. Available: https://www.noaa.gov/blue-economy R. Thandiackal and G. V. Lauder, “How zebrafish turn: analysis of pressure force dynamics and mechanical work,” J. Exp. Biol. , vol. 223, no. 16, p. jeb223230, Aug. 2020, doi: 10.1242/jeb.223230. F. E. Fish and J. T. Beneski, “Evolution and Bio-Inspired Design: Natural Limitations,” in Biologically Inspired Design: Computational Methods and Tools , A. K. Goel, D. A. McAdams, and R. B. Stone, Eds., London: Springer, 2014, pp. 287–312. doi: 10.1007/978-1-4471-5248-4_12. P. R. Bandyopadhyay, “Trends in biorobotic autonomous undersea vehicles,” IEEE J. Ocean. Eng. , vol. 30, no. 1, pp. 109–139, Jan. 2005, doi: 10.1109/JOE.2005.843748. F. E. Fish and G. V. Lauder, “Control surfaces of aquatic vertebrates: active and passive design and function,” J. Exp. Biol. , vol. 220, no. 23, pp. 4351–4363, Dec. 2017, doi: 10.1242/jeb.149617. G. Li, G. Liu, D. Leng, X. Fang, G. Li, and W. Wang, “Underwater undulating propulsion biomimetic robots: a review,” Biomimetics , vol. 8, no. 3, Art. no. 3, Jul. 2023, doi: 10.3390/biomimetics8030318. P. R. Bandyopadhyay, “Guest Editorial: Biology-Inspired Science and Technology for Autonomous Underwater Vehicles,” IEEE J. Ocean. Eng. , vol. 29, no. 3, pp. 542–546, Jul. 2004, doi: 10.1109/JOE.2004.833099. R. Salazar, V. Fuentes, and A. Abdelkefi, “Classification of biological and bioinspired aquatic systems: A review,” Ocean Eng. , vol. 148, pp. 75–114, Jan. 2018, doi: 10.1016/j.oceaneng.2017.11.012. A. Wm. English, “Limb movements and locomotor function in the California sea lion (Zalophus californianus),” J. Zool. , vol. 178, no. 3, pp. 341–364, 1976, doi: 10.1111/j.1469-7998.1976.tb02274.x. S. D. Feldkamp, “Swimming in the California sea lion: morphometrics, drag and energetics,” J. Exp. Biol. , vol. 131, no. 1, pp. 117–135, Sep. 1987, doi: 10.1242/jeb.131.1.117. S. D. Feldkamp, “Foreflipper propulsion in the California sea lion, Zalophus californianus,” J. Zool. , vol. 212, no. 1, pp. 43–57, 1987, doi: 10.1111/j.1469-7998.1987.tb05113.x. C. Friedman and M. C. Leftwich, “The kinematics of the California sea lion foreflipper during forward swimming,” Bioinspir. Biomim. , vol. 9, no. 4, p. 046010, Nov. 2014, doi: 10.1088/1748-3182/9/4/046010. S. D. Feldkamp, R. L. DeLong, and G. A. Antonelis, “Diving patterns of California sea lions, Zalophus californianus,” Can. J. Zool. , vol. 67, no. 4, pp. 872–883, Apr. 1989, doi: 10.1139/z89-129. F. E. Fish, “Speed,” in Encyclopedia of Marine Mammals , W.F. Perrin, B. Wursig, and J.G.M. Thewissen., Academic Press, San Diego, 2002, pp. 1161–1163. D. K. Odell, “California sea lion Zalophus californianus,” in Handbook of Marine Mammals , S.H. Ridgway and R. Harrison., vol. 1, Academic Press. London, 1981, pp. 67–97. S. Godfrey, “Additional observations of subaqueous locomotion in the California Sea Lion (Zalophus californianus),” Aquat. Mamm. , vol. 11, no. 2, pp. 53–57, 1985. F. E. Fish, J. Hurley, and D. P. Costa, “Maneuverability by the sea lion Zalophus californianus: turning performance of an unstable body design,” J. Exp. Biol. , vol. 206, no. 4, pp. 667–674, 2003, doi: 10.1242/jeb.00144. S. Kadapa, A. Drago, N. Marcouiller, J. L. Tangorra, and H. G. Kwatny, “Development of numerical model for a bio-inspired sea lion robot,” IEEE J. Ocean. Eng. , 18 2025. N. M. Puzai, A. F. Ayob, and M. R. Arshad, “A review on recent advancements in unmanned underwater vehicle design,” J Ocean Mech Aerosp.-Sci Eng , vol. 31, no. 1, pp. 1–8, 2016. M.-G. Kim, H. Kang, M.-J. Lee, G. R. Cho, J.-H. Li, and C. Kim, “UUV platform optimal design for overcoming strong current,” J. Ocean Eng. Technol. , vol. 35, no. 6, pp. 434–445, Dec. 2021, doi: 10.26748/KSOE.2021.069. “Underwater Gliders for Ocean Research,” ResearchGate. Accessed: Jan. 23, 2025. [Online]. Available: https://www.researchgate.net/publication/233687927_Underwater_Gliders_for_Ocean_Research F. E. Fish, “Advantages of aquatic animals as models for bio-inspired drones over present AUV technology,” Bioinspir. Biomim. , vol. 15, no. 2, p. 025001, Feb. 2020, doi: 10.1088/1748-3190/ab5a34. A. P. Mignano, S. Kadapa, A. C. Drago, G. V. Lauder, H. G. Kwatny, and J. L. Tangorra, “Fish robotics: multi-fin propulsion and the coupling of fin phase, spacing, and compliance,” Bioinspir. Biomim. , vol. 19, no. 2, p. 026006, Jan. 2024, doi: 10.1088/1748-3190/ad1dba. A. P. Mignano, S. Kadapa, J. L. Tangorra, and G. V. Lauder, “Passing the Wake: Using Multiple Fins to Shape Forces for Swimming,” Biomimetics , vol. 4, no. 1, 2019, doi: 10.3390/biomimetics4010023. C. H. White, G. V. Lauder, and H. Bart-Smith, “Tunabot Flex: a tuna-inspired robot with body flexibility improves high-performance swimming,” Bioinspir. Biomim. , vol. 16, no. 2, p. 026019, Mar. 2021, doi: 10.1088/1748-3190/abb86d. R. Baines et al. , “Multi-environment robotic transitions through adaptive morphogenesis,” Nature , vol. 610, no. 7931, pp. 283–289, Oct. 2022, doi: 10.1038/s41586-022-05188-w. Z. Chen, T. I. Um, J. Zhu, and H. Bart-Smith, “Bio-Inspired Robotic Cownose Ray Propelled by Electroactive Polymer Pectoral Fin,” in Volume 2: Biomedical and Biotechnology Engineering; Nanoengineering for Medicine and Biology , Denver, Colorado, USA: ASMEDC, Jan. 2011, pp. 817–824. doi: 10.1115/IMECE2011-64174. P. Liljebäck and R. Mills, “Eelume: A flexible and subsea resident IMR vehicle,” in OCEANS 2017 - Aberdeen , Jun. 2017, pp. 1–4. doi: 10.1109/OCEANSE.2017.8084826. E. Kelasidi, P. Liljeback, K. Y. Pettersen, and J. T. Gravdahl, “Innovation in Underwater Robots: Biologically Inspired Swimming Snake Robots,” IEEE Robot. Autom. Mag. , vol. 23, no. 1, pp. 44–62, Mar. 2016, doi: 10.1109/MRA.2015.2506121. S. Randeni, E. M. Mellin, M. Sacarny, S. Cheung, M. Benjamin, and M. Triantafyllou, “Bioinspired morphing fins to provide optimal maneuverability, stability, and response to turbulence in rigid hull AUVs,” Bioinspir. Biomim. , vol. 17, no. 3, p. 036012, Apr. 2022, doi: 10.1088/1748-3190/ac5a3d. Y. Wang, Y. Guo, S. Yang, T. Sun, X. Wang, and H. Zhou, “Design, Hydrodynamic Analysis, and Testing of a Bio-inspired Movable Bow Mechanism for the Hybrid-driven Underwater Glider,” J. Bionic Eng. , vol. 20, no. 4, pp. 1493–1513, Jul. 2023, doi: 10.1007/s42235-023-00361-x. A. M. Leahy et al. , “The role of California sea lion (Zalophus californianus) hindflippers as aquatic control surfaces for maneuverability,” J. Exp. Biol. , vol. 224, no. 20, p. jeb243020, Oct. 2021, doi: 10.1242/jeb.243020. S. J. Kerr, F. E. Fish, A. J. Nicastro, J. A. Zeligs, S. Skrovan, and M. C. Leftwich, “Biomechanical energetics of terrestrial locomotion in California sea lions (Zalophus californianus),” J. Exp. Biol. , vol. 225, no. 18, p. jeb244163, Sep. 2022, doi: 10.1242/jeb.244163. G. Perrotta, F. E. Fish, D. S. Adams, A. M. Leahy, A. M. Downs, and M. C. Leftwich, “Velocity Field Measurements of the California Sea Lion Propulsive Stroke Using Bubble PIV,” Fluids , vol. 7, no. 1, Art. no. 1, Jan. 2022, doi: 10.3390/fluids7010003. P. W. Webb, “Maneuverability - general issues,” IEEE J. Ocean. Eng. , vol. 29, no. 3, pp. 547–555, Jul. 2004, doi: 10.1109/JOE.2004.833220. A. Drago, S. Kadapa, N. Marcouiller, H. G. Kwatny, and J. L. Tangorra, “Using Reinforcement Learning to Develop a Novel Gait for a Bio-Robotic California Sea Lion,” Biomimetics , vol. 9, no. 9, p. 522, Aug. 2024, doi: 10.3390/biomimetics9090522. Additional Declarations No competing interests reported. Supplementary Files DevelopmentofaBioroboticSwimmerBasedontheCaliforniaSeaLionCompressed.mp4 Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7455024","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":511841063,"identity":"23116525-8780-4f23-a039-b0b82e67cb2d","order_by":0,"name":"Nicholas Marcouiller","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwElEQVRIiWNgGAWjYFAC5uO/f1Qwg1iMB4A8YrSwJUgznIGoJFYLj4E0YxspWszFjiUYF86zzueXSD5wgKHCOrGBkBbL2ckHkmduS7ec2XMs4QDDmXTCWgxupyUc4N122MDgeI/BAca2w8RoyTFs4J1z2MD+MP+HA4z/iNNizMzbALSFvYfhAGMDUVrS0hhnHEs3kDhzzOBAwrF0YyK0JB9j+FBjbcA/I/nhAyBDlqAWVJBAmvJRMApGwSgYBbgAAPToQxxaWt4GAAAAAElFTkSuQmCC","orcid":"","institution":"Drexel University","correspondingAuthor":true,"prefix":"","firstName":"Nicholas","middleName":"","lastName":"Marcouiller","suffix":""},{"id":511841064,"identity":"868996f6-b897-421a-9e37-e9d158130dea","order_by":1,"name":"Shraman Kadapa","email":"","orcid":"","institution":"Drexel University","correspondingAuthor":false,"prefix":"","firstName":"Shraman","middleName":"","lastName":"Kadapa","suffix":""},{"id":511841065,"identity":"e4e3851a-2878-433f-8ce4-b14213c0c770","order_by":2,"name":"Anthony Drago","email":"","orcid":"","institution":"Drexel University","correspondingAuthor":false,"prefix":"","firstName":"Anthony","middleName":"","lastName":"Drago","suffix":""},{"id":511841067,"identity":"be15185d-d77a-427b-a67f-321e0ed8d983","order_by":3,"name":"Frank Fish","email":"","orcid":"","institution":"West Chester University","correspondingAuthor":false,"prefix":"","firstName":"Frank","middleName":"","lastName":"Fish","suffix":""},{"id":511841068,"identity":"65b3a67b-3204-4d9d-96a9-587de6893de6","order_by":4,"name":"Megan Leftwich","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Megan","middleName":"","lastName":"Leftwich","suffix":""},{"id":511841069,"identity":"535ad03e-1d09-4899-b904-61bf7b0ce778","order_by":5,"name":"Harry Kwatny","email":"","orcid":"","institution":"Drexel University","correspondingAuthor":false,"prefix":"","firstName":"Harry","middleName":"","lastName":"Kwatny","suffix":""},{"id":511841070,"identity":"d6dd572c-5e3a-43b0-a5bf-49a04368b263","order_by":6,"name":"James Tangorra","email":"","orcid":"","institution":"Drexel University","correspondingAuthor":false,"prefix":"","firstName":"James","middleName":"","lastName":"Tangorra","suffix":""}],"badges":[],"createdAt":"2025-08-25 15:08:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7455024/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7455024/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":91078183,"identity":"2b7a1bf7-b55b-475e-aacd-8ab9781b52ff","added_by":"auto","created_at":"2025-09-11 11:18:38","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":430198,"visible":true,"origin":"","legend":"\u003cp\u003eCalifornia sea lion and SEAMOUR performing banking turn maneuver\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/139fe71ec33a7c9507044866.png"},{"id":91074275,"identity":"44b089cf-584b-45e2-89bb-3152305e4e30","added_by":"auto","created_at":"2025-09-11 11:02:39","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":625241,"visible":true,"origin":"","legend":"\u003cp\u003eCalifornia sea lion during various stages of swimming and maneuvering. (a) Mid propulsive stroke, (b) Gliding, (c) Banking turn, (d) Back flip, (e) Darting to track object\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/65cbff3c4628892a71355f03.jpeg"},{"id":91074274,"identity":"940db9c4-5b54-4c61-9391-1ee93c0e49fe","added_by":"auto","created_at":"2025-09-11 11:02:38","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":277727,"visible":true,"origin":"","legend":"\u003cp\u003eCharacteristic stroke recovery, power, and paddle phases from both side and bottom views.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/2b14b983c3dbe221946f6e46.png"},{"id":91074251,"identity":"7467a075-0c03-4f40-8b1f-9f89d3827a11","added_by":"auto","created_at":"2025-09-11 11:02:38","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":360397,"visible":true,"origin":"","legend":"\u003cp\u003eCalifornia sea lion performing a 90° banking turn. (a) Before turn. (b) During turn. (c) Coming out of turn. (d) After turn.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/1ab0433e5bcd84342bf94a15.png"},{"id":91074278,"identity":"46be3ebf-c9d2-469c-813c-3477f4c2b256","added_by":"auto","created_at":"2025-09-11 11:02:39","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":189476,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Sea lion body frame of reference. (b) Areas of interest- head (red), fore flippers (blue), pelvis (green), and hind flippers (purple). (c) Section of sea lion fore flipper\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/bdc5c51a61b29721948a9166.png"},{"id":91074250,"identity":"3bb9a836-f9ba-4d20-93a0-ba36855981e9","added_by":"auto","created_at":"2025-09-11 11:02:38","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":343847,"visible":true,"origin":"","legend":"\u003cp\u003eKinovea Tracking software used for decomposition of fore flipper kinematics. Rotations taken in flipper frame. (Left) Pitch\u003csub\u003ef\u003c/sub\u003e, (Center) Yaw\u003csub\u003ef\u003c/sub\u003e, (Right) Roll\u003csub\u003ef\u003c/sub\u003e\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/5847af8789210a2ae668ddf1.png"},{"id":91076509,"identity":"5ed4640b-8267-401e-ba61-6a7f6a51ab13","added_by":"auto","created_at":"2025-09-11 11:10:39","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":656360,"visible":true,"origin":"","legend":"\u003cp\u003eFore flipper frame of reference. Yellow axis longitudinal through center of flipper represents roll\u003csub\u003ef\u003c/sub\u003e, green axis perpendicular to flipper at shoulder represents yaw\u003csub\u003ef\u003c/sub\u003e, red axis parallel to body at shoulder joint represents pitch\u003csub\u003ef\u003c/sub\u003e\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/e0ef8563a8b922fc4f7aa089.png"},{"id":91074263,"identity":"cfb9161d-b0b0-4cad-add1-6b906b3d25f4","added_by":"auto","created_at":"2025-09-11 11:02:38","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":274106,"visible":true,"origin":"","legend":"\u003cp\u003eTop: SEAMOUR with head pitched 60°, fore flippers yawed\u003csub\u003ef\u003c/sub\u003e and rolled\u003csub\u003ef\u003c/sub\u003e 90°, pelvis pitched 60°, hind flippers yawed 25° and rolled 0°.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBottom: SEAMOUR in streamlined position. (a) Head- pitch and yaw DOF (b) Fore flipper motor configuration- rollf, yaw\u003csub\u003ef\u003c/sub\u003e, and pitch\u003csub\u003ef\u003c/sub\u003e DOF (c) Waterproof power and electronics box (d) Pelvis- pitch and yaw DOF (e) Hind flippers- yaw and roll DOF\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/44ce81ae8712f5cde5bca509.png"},{"id":91074269,"identity":"01191dbd-1f91-441e-8baf-90cdc919c2ce","added_by":"auto","created_at":"2025-09-11 11:02:38","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":531008,"visible":true,"origin":"","legend":"\u003cp\u003eLocation of waterproof box for electronics and schematic of electronics housed inside box.\u003c/p\u003e","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/872bb8020fb47abecfd72d0c.png"},{"id":91074252,"identity":"531ffd06-8fb4-452f-9877-f7514a1cdd57","added_by":"auto","created_at":"2025-09-11 11:02:38","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":281431,"visible":true,"origin":"","legend":"\u003cp\u003eSEAMOUR mounted on unconstrained docking system (purple box) for straight swimming and maneuvering experiments\u003c/p\u003e","description":"","filename":"floatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/b98c06ca80917600980f54d2.png"},{"id":91074262,"identity":"d7ae3c2e-951e-4a58-8397-db7a83e3ae23","added_by":"auto","created_at":"2025-09-11 11:02:38","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":855964,"visible":true,"origin":"","legend":"\u003cp\u003eFore flipper motor assembly. (a) Pitch\u003csub\u003ef\u003c/sub\u003e motor. (b) Yaw\u003csub\u003ef\u003c/sub\u003e motor. (c) Roll\u003csub\u003ef\u003c/sub\u003e motor.\u003c/p\u003e","description":"","filename":"floatimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/a7c478008ddabccc9addd785.png"},{"id":91078184,"identity":"54eade9f-0676-4b36-90f4-d11d96507a54","added_by":"auto","created_at":"2025-09-11 11:18:38","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":488246,"visible":true,"origin":"","legend":"\u003cp\u003eEvaluation and validation of fore flipper design. Left: Biological flipper shape and bending. Center: Fore flipper support structure during FEA. Right: Fabricated silicone cased fore flipper during experimental bending validation.\u003c/p\u003e","description":"","filename":"floatimage12.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/a7773dc29447c5fef7728f1a.png"},{"id":91074289,"identity":"13a36b1e-f917-4d70-8d89-ebb0b6924ed5","added_by":"auto","created_at":"2025-09-11 11:02:39","extension":"png","order_by":13,"title":"Figure 13","display":"","copyAsset":false,"role":"figure","size":414053,"visible":true,"origin":"","legend":"\u003cp\u003eFlipper casting process, depicted using fore flippers. Left to right: FDM 3D printed ABS support structure, support structure mounted in casting mold, mold assembled for casting, casted fore flipper\u003c/p\u003e","description":"","filename":"floatimage13.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/d9cfe9a625ba0a28da5bc2ee.png"},{"id":91076522,"identity":"32c1bf27-cc8c-4b13-aca6-a1aa9674e582","added_by":"auto","created_at":"2025-09-11 11:10:40","extension":"png","order_by":14,"title":"Figure 14","display":"","copyAsset":false,"role":"figure","size":387356,"visible":true,"origin":"","legend":"\u003cp\u003eHead yawed 60° without (a1) and with (a2) four-way stretch fabric. (b) Pelvic section with exposed hind flipper yaw and roll motors.\u003c/p\u003e","description":"","filename":"floatimage14.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/44bae34156e9fe007e7583e6.png"},{"id":91074322,"identity":"663d9b71-c68c-4a66-b2e9-5c603987d77b","added_by":"auto","created_at":"2025-09-11 11:02:40","extension":"png","order_by":15,"title":"Figure 15","display":"","copyAsset":false,"role":"figure","size":156351,"visible":true,"origin":"","legend":"\u003cp\u003e(a) SEAMOUR body frame of reference plane and axis in CAD. (b) Center of gravity (purple circle) and center of buoyancy region (orange box) for the entire system\u003c/p\u003e","description":"","filename":"floatimage15.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/27c0db4d06590e2dce425d26.png"},{"id":91074259,"identity":"849754d7-b31b-4cc2-9b16-31ab9133b12b","added_by":"auto","created_at":"2025-09-11 11:02:38","extension":"png","order_by":16,"title":"Figure 16","display":"","copyAsset":false,"role":"figure","size":93605,"visible":true,"origin":"","legend":"\u003cp\u003eBody orientation for each static stability test. Orange arrow indicates force from COB and purple arrow is force from COG. Rotation due to restoring moment depicted by direction of green arrow. (a) Roll test with fore flippers in streamlined position. (b) Roll test with fore flippers in extended position (yawed\u003csub\u003ef\u003c/sub\u003e 90° and rolled\u003csub\u003ef\u003c/sub\u003e 90°). (c) Pitch test with fore flippers in streamlined position.\u0026nbsp;\u003c/p\u003e","description":"","filename":"floatimage16.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/1bdddff5aaa878b03002e829.png"},{"id":91076511,"identity":"930e5761-9fe9-4778-b414-7f7c066dfe92","added_by":"auto","created_at":"2025-09-11 11:10:39","extension":"jpeg","order_by":17,"title":"Figure 17","display":"","copyAsset":false,"role":"figure","size":491898,"visible":true,"origin":"","legend":"\u003cp\u003eTime sequence from t = 0 to t = 2 seconds showing the three swimming techniques implemented, represented with a CAD model. The final second of each stroke, corresponding to the recovery phase when the flippers reset for the next stroke, is omitted to better highlight the propulsive phases. Characteristic Stroke (Top). Characteristic stroke with pelvis actuation (Middle). Modified side paddle stroke (Bottom).\u003c/p\u003e","description":"","filename":"floatimage17.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/610d461bde6bd8f473c8e4f6.jpeg"},{"id":91076516,"identity":"002168d8-489d-45db-b82f-29fc37aae60e","added_by":"auto","created_at":"2025-09-11 11:10:39","extension":"png","order_by":18,"title":"Figure 18","display":"","copyAsset":false,"role":"figure","size":159619,"visible":true,"origin":"","legend":"\u003cp\u003eRoll stability plots for initial angles of 20°, 40°, 60°, and 80°. The black dot on each line represents the settling time. (a) time response with fore flippers yawed\u003csub\u003ef\u003c/sub\u003e 0° and (b) time response for flippers yawed\u003csub\u003ef\u003c/sub\u003e \u0026nbsp;and rolled\u003csub\u003ef\u003c/sub\u003e \u0026nbsp;90°.\u0026nbsp;\u003c/p\u003e","description":"","filename":"floatimage18.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/2f9c28dcdb4c97925791df2b.png"},{"id":91074318,"identity":"d037a9aa-381c-40a3-8789-2cfe225513e4","added_by":"auto","created_at":"2025-09-11 11:02:40","extension":"png","order_by":19,"title":"Figure 19","display":"","copyAsset":false,"role":"figure","size":160778,"visible":true,"origin":"","legend":"\u003cp\u003ePitch stability time response plot for 20°, 40°, 60°, and 80°. The black dot on each line represents the settling time.\u003c/p\u003e","description":"","filename":"floatimage19.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/3ee12402b0ea8d356e8c8b89.png"},{"id":91076496,"identity":"b2db32cf-7752-416b-8ba5-82f670c02bd8","added_by":"auto","created_at":"2025-09-11 11:10:38","extension":"png","order_by":20,"title":"Figure 20","display":"","copyAsset":false,"role":"figure","size":257839,"visible":true,"origin":"","legend":"\u003cp\u003e(Left) Position plot for three different stoke types. (Right) Pitch angle for three different stoke types.\u003c/p\u003e","description":"","filename":"floatimage20.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/0874a35242231f6a698d8016.png"},{"id":91074280,"identity":"84942623-60d6-4f6b-910f-aeb15ec7fac2","added_by":"auto","created_at":"2025-09-11 11:02:39","extension":"png","order_by":21,"title":"Figure 21","display":"","copyAsset":false,"role":"figure","size":266802,"visible":true,"origin":"","legend":"\u003cp\u003eSwimming Trials: SEAMOUR during fourth stroke at t=10s\u003c/p\u003e","description":"","filename":"floatimage21.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/439a09289b88ef9c7027ae34.png"},{"id":91074267,"identity":"dd278b11-a0d1-4108-bfcd-7c8cd0c34cca","added_by":"auto","created_at":"2025-09-11 11:02:38","extension":"png","order_by":22,"title":"Figure 22","display":"","copyAsset":false,"role":"figure","size":263413,"visible":true,"origin":"","legend":"\u003cp\u003eHead, pelvic, and combined actuation for pitch and yaw maneuvers\u003c/p\u003e","description":"","filename":"floatimage22.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/e2f6fc0a0eae11150d667e3d.png"},{"id":91078186,"identity":"fdb9a56e-a09a-4761-bd57-671e947eb8eb","added_by":"auto","created_at":"2025-09-11 11:18:39","extension":"png","order_by":23,"title":"Figure 23","display":"","copyAsset":false,"role":"figure","size":250843,"visible":true,"origin":"","legend":"\u003cp\u003e(Left) Head and pelvis articulated 30° during yaw test at t=3s. (Right) Head and pelvis articulated 60° during pitch test at t=3s.\u003c/p\u003e","description":"","filename":"floatimage23.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/b6d4c464f0d531991705c09b.png"},{"id":91074256,"identity":"0c5e31ab-ea29-4c54-9224-906bcd319267","added_by":"auto","created_at":"2025-09-11 11:02:38","extension":"png","order_by":24,"title":"Figure 24","display":"","copyAsset":false,"role":"figure","size":811937,"visible":true,"origin":"","legend":"\u003cp\u003eSEAMOUR executing curvilinear path using coordinated flipper and body motions. Images overlayed from 4s to 24s at intervals of 4s.\u003c/p\u003e","description":"","filename":"floatimage24.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/c25dc94bb4a6140f2063f4e5.png"},{"id":91074310,"identity":"329d7217-8c88-4ce5-8d19-dc5ab284987d","added_by":"auto","created_at":"2025-09-11 11:02:40","extension":"png","order_by":25,"title":"Figure 25","display":"","copyAsset":false,"role":"figure","size":409679,"visible":true,"origin":"","legend":"\u003cp\u003eCoordinated maneuvers. Still images taken to show distinct portions of the maneuver at key, non-uniform intervals. (a) Single flipper flapped to roll the body. (b) Banking turn using yaw, roll, and pitch rotations. (c) Front flip turn.\u003c/p\u003e","description":"","filename":"floatimage25.png","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/adc5e9d4417be3018ad6beb6.png"},{"id":101397729,"identity":"8b64300a-bb41-481b-87d5-4fae86ee92bb","added_by":"auto","created_at":"2026-01-29 09:36:09","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":12659602,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/112c0dd5-5d6f-4a72-a053-689e46a69ca8.pdf"},{"id":101008793,"identity":"ccf2186f-c983-40a6-9fc0-55c6feb85dbc","added_by":"auto","created_at":"2026-01-23 18:42:48","extension":"mp4","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":19858433,"visible":true,"origin":"","legend":"","description":"","filename":"DevelopmentofaBioroboticSwimmerBasedontheCaliforniaSeaLionCompressed.mp4","url":"https://assets-eu.researchsquare.com/files/rs-7455024/v1/c516be6110f497ca0bc1407c.mp4"}],"financialInterests":"No competing interests reported.","formattedTitle":"Development of a Bio-robotic Swimmer Based on the California Sea Lion","fulltext":[{"header":"1 Introduction ","content":"\u003cp\u003eUnmanned Underwater Vehicles (UUVs) play a crucial role across a wide array of maritime operations, including scientific research, commercial endeavors, and defense applications; however, their limited maneuverability often restricts their operational reach. While UUVs are increasingly used in the open ocean, there is growing interest in expanding their capabilities to littoral zones, rivers, harbors, and surf zones where maneuverability plays a key role in their success [1], [2]. These missions would frequently require UUVs to navigate tight spaces, avoid obstacles, and execute rapid turns and directional changes in response to turbulent and occluded flow conditions. To operate effectively in these complex, high-energy environments, UUVs would need to be highly agile and maneuverable, requiring further advancements of UUVs.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAquatic animals, particularly those adept at navigating high-energy environments, serve as valuable models for enhancing the maneuverability of underwater vehicles. Their remarkable locomotion strategies, which include coordinated body segment movement and adaptive high-speed swimming patterns, enable agile maneuvering in complex, cluttered environments [3]. Studying the morphology, physiology, and kinematics of these biological systems directly informs the development of robotic platforms with improved propulsive efficiency, maneuverability, agility, stability, station holding, and stealth [4]. While significant progress has been made in understanding underwater animal locomotion [5], [6], current underwater systems have yet to fully exploit the characteristics such as body-fin coordination, adaptable compliance, or dynamic vortex manipulation that enable certain aquatic animals to maneuver effectively in diverse flow conditions [7], [8], [9]. \u0026nbsp;Among these, the California sea lion (\u003cem\u003eZalophus californianus\u003c/em\u003e) stands out as a prime subject of advanced research [10], [11], [12], [13]. The California sea lion is capable of swimming at 9.7 m/s and diving to a maximum depth of 274 m \u0026nbsp;[14], [15], [16]. Its exceptional ability to navigate complex underwater environments, where it encounters obstacles and energetic flows, makes it an invaluable model for a bio-inspired design [17], [18]. Notably, sea lions have achieved turning rates of up to 690\u0026deg;/s and minimum unpowered (i.e., without active propulsion) turning radii as small as 0.09 body lengths [18]. This high degree of maneuverability and agility can be partly attributed to the animal\u0026apos;s coordinated use of its powerful flippers and flexible body [10], [17]. Ultimately, understanding these animals\u0026apos; swimming and maneuvering strategies provides a crucial foundation for applying such techniques in dynamic conditions, offering key insights for developing UUVs capable of operating across varied flow regimes.\u003c/p\u003e\n\u003cp\u003eThe objective of this work was to develop a freely swimming bio-robotic research platform, modeled after the California sea lion, to explore morphological and kinematic traits that contribute to effective swimming and maneuvering. This platform, named the Stroke Experimentation and Maneuver Optimizing Underwater Robot (SEAMOUR) (Figure 1), was specifically designed to investigate various propulsion and maneuvering techniques through coordinated flipper and body motions. Initial tests explored rectilinear swimming approaches and the use of multiple control surfaces\u0026mdash;located fore and aft of the robot\u0026apos;s center of mass\u0026mdash;to produce pitch and yaw turns. Insights from these studies enabled SEAMOUR to execute freely swimming, coordinated multi-axis maneuvers using independently actuated control surfaces distributed along its body. The system\u0026apos;s repeatable and versatile functionality makes it ideal for understanding the diverse swimming and maneuvering techniques employed by California sea lions. A complementary numerical model of the bio-robotic platform was also developed to evaluate and explore various swimming strategies; however, its details are presented in a separate study [19] and are not the focus of this paper. Studies using both the physical and numerical systems provide a foundation for identifying design features that can enhance UUV maneuverability, with future work exploring performance in dynamic flow environments.\u003c/p\u003e\n\u003cp\u003eWhile traditional UUVs possess varied\u0026nbsp;strengths, their inherent characteristics consistently limit their effectiveness in complex, high-energy environments. For example, torpedo-shaped UUVs that use a single axial propeller operate with limited control surfaces and are designed to prioritize stability and speed for navigating unobstructed open water, an environment where they consistently perform well [20]. While box-frame UUVs can be equipped with a variety of payloads and have multiple propulsors, their large, flat shape generates significant drag at high speeds and can be easily perturbed in energetic environments [21]. Lastly, glider-type UUVs are highly efficient for long-distance travel in open water but have limited speed and agility [22]. As a result, there remains a critical need for systems capable of operating in dynamic coastal environments. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBiomimetic robotics has demonstrated significant success in developing versatile systems that leverage intricate biological traits, offering a powerful alternative to conventional engineering paradigms. While propeller-based systems allow for precise and predictable adjustments to speed and direction through direct control of rotation speed and thrust, animals, in contrast, use intricate movements of their flexible appendages and bodies to generate multidimensional forces [23], [18]. Over the years, numerous researchers have worked on developing systems inspired by marine creatures, including various fish to understand the use of multiple fins for propulsion and maneuvering [24], [25], [26]; sea turtles and cownose rays to explore the potential of soft actuators [27], [28]; and sea snakes/eels to investigate the use of multi-bodied systems [29], [30], to name a few. By incorporating biologically derived characteristics into traditionally shaped UUVs, some researchers have already demonstrated improvements in system maneuverability [31], [32]. The application of biological principles consistently underscores the immense potential of biomimetics, further validating the approach of using highly agile models like the California sea lion to significantly advance UUV maneuverability and performance across diverse environments.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe remainder of this paper is organized as follows. The first section provides a description of common techniques used by sea lions for swimming and maneuvering, followed by the extraction of flipper and body kinematics. The next section describes the design of the bio-robotic system, with the following section describing the experimental setup and data collection techniques. A combined results and discussion section examine the static stability of the system\u0026rsquo;s design, rectilinear swimming approaches, pitch and yaw turns, and coordinated maneuvers. Finally, the paper will conclude with a summary of key findings of the work and a discussion of future research directions.\u003c/p\u003e"},{"header":"2 Background of California Sea Lion Swimming and Maneuvering","content":"\u003cp\u003eThe California sea lion (\u003cem\u003eZalophus californianus\u003c/em\u003e) was selected as the model organism for this study due to its versatility and exceptional performance in both open ocean and dynamic coastal environments. Sea lions regularly navigate complex habitats including kelp forests, rocky shorelines, and turbulent surf zones, making maneuverability a critical factor in their ecological success [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Adult sea lions, which typically measure around 1.81 m in length, exhibit impressive aquatic performance\u0026mdash;cruising at sustained speeds of 3.5 m/s [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], executing porpoising leaps of 0.94 m in height and 1.90 m in length at 2 m/s, and performing agile banking turns with minimum radii of 0.29 m at 3.2 m/s [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Their capabilities also extend beyond the water, as they can transition to land and move efficiently using walking or galloping gaits [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Sea lions retain mobility in both the fore and hind limbs, providing distinct mechanical strategies for propulsion, stabilization, and maneuvering. In the water, the fore flippers serve as the primary propulsors and control surfaces, while the hind flippers assist with stabilization and maneuvering [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Together, these traits make the California sea lion an ideal biological model for investigating aquatic locomotion, offering high maneuverability through coordinated use of its flippers and body across a range of aquatic environments (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe California sea lion primarily relies on its fore flippers for propulsion during swimming, with minimal contribution from the hind flippers or other body sections. The propulsive stroke has been categorized into three distinct phases: recovery, power, and paddle (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], defined by the motion at the base of the flipper (shoulder/elbow joints). The recovery phase begins and ends with the fore flippers retracted, medially rotated, and held against the ventrolateral surface of the body. During the recovery phase, the flippers are laterally rotated and abducted, bringing them near a zero angle of attack with the flow. As the flippers are further abducted, they rotate to a positive angle of attack. Lateral rotation slows as medial rotation begins. The power phase starts as medial rotation continues, and the flippers are adducted, resulting in a noticeable dorsal flexion of the distal two-thirds. The flippers' orientation then changes from a positive to a negative angle of attack as they pass the body midline. Rotation slows as adduction continues until the flippers are fully extended beneath the body. At the start of the paddle phase of the stroke, the flippers are maximally adducted and medial rotation continues as they are retracted. The flippers are then forcefully brought upward and inward toward the body [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Finally, the flippers return to a streamlined position next to the body [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. In steady swimming, the hind flippers are typically positioned in an inverted V-shape with little head movement [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] and between strokes the fore flippers may be adducted against the body to create a streamlined profile [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Throughout these studies this stroke is referred to as the characteristic stroke and acts as the foundation to other swimming approaches that are explored using the bio-robotic system.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe California sea lion employs a variety of strategies to maneuver through the water, with one of the most common being the execution of a banking turn. This maneuver showcases the animal\u0026rsquo;s high maneuverability and highlights the coordinated use of body and flipper movements required to rotate the body about multiple axes. Notably, banking turns enable sharp, unpowered changes in direction, achieving smaller turning radii than powered turns [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. To initiate a banking turn, the sea lion first rotates its head laterally toward the desired direction of travel. This motion is followed by abduction of the fore flippers and axial body rotation about the longitudinal axis. As the body continues to roll, the hind flippers are angled dorsally into the vertical plane, with an abduction of the digits. Simultaneously, the body undergoes lateral flexion as the head and pelvic region bend toward the center of the turn as the body is bent dorsally producing a continuous, coordinated turn. To complete the turn the animal straightens its body, reducing flexion of the vertebral column and re-aligning with the intended direction of travel. At the same time, the fore flippers are used to accelerate the body out of the turn as the flippers are adducted and retracted to lay appressed against the body. The body then reverses its roll, returning to its starting orientation to resume steady swimming or gliding (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. The coordinated body and flipper kinematics demonstrated in this maneuver provide valuable biomechanical principles for informing the design of the robotic system and guiding the development of maneuverability strategies.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"3 Kinematics of Biological System","content":"\u003cp\u003eThe kinematics of flipper and body motions during swimming and maneuvering were identified from videos of both trained and untrained California sea lions. Footage was captured at the Smithsonian Zoological Park (Washington DC) using a stationary high-definition camera (GC-PX100BU, JVC, Japan) recording at 60 frames per second [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. The large underwater viewing window allowed for observation over a distance of at least three body lengths as the sea lions swam passively. Additional recordings of trained sea lions performing directed maneuvers were collected at SeaWorld Florida and SLEWTHS at Moss Landing Marine Laboratories [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. From these observations, four features were identified: a manipulable head (cervical), an articulatable pelvis (lumbar), hind flippers capable of acting as control surfaces, and fore flippers that serve both as primary propulsion and control surfaces (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTo inform the design of robotic platforms, the swimming and maneuvering kinematics of California sea lions were analyzed, specifically focusing on the movements of their flippers and various body segments. The analysis was done using the two-dimensional digital tracking software Kinovea (Kinovea.org, France), which was used to measure rotations about different axes for the fore flippers, hind flippers, pelvis, and head. Video footage was selected where sea lions swam either parallel or perpendicular to the viewing window, ensuring clear visualization of specific movements. Kinematics were identified by analyzing videos that provided an unobstructed view of the targeted body regions and their respective angles (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). Since each motion was captured from a single perspective, multiple videos from different angles were used to reconstruct a complete movement profile. The data was then aggregated and averaged to establish a standard motion, providing insight into the degrees of freedom required for each body segment. These findings directly informed the design of the actuators for the robotic platform.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTo model the 3D kinematics of fore flipper motion, an analysis was done that focused on the rotation at the base of the flipper during the characteristic propulsive stroke (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). The motion was categorized into the three phases, involving roll\u003csub\u003ef\u003c/sub\u003e, yaw\u003csub\u003ef\u003c/sub\u003e, and pitch\u003csub\u003ef\u003c/sub\u003e rotations within the flipper\u0026rsquo;s frame of reference (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Throughout the stroke, the flippers moved through 140\u0026deg; of roll\u003csub\u003ef\u003c/sub\u003e, 110\u0026deg; of yaw\u003csub\u003ef\u003c/sub\u003e, and 110\u0026deg; of pitch \u003csub\u003ef\u003c/sub\u003e. The rotations during these motions were plotted, and key inflection points were identified as control points. To validate the stroke analysis, a Piecewise Cubic Hermite Interpolating Polynomial (PCHIP) spline was fitted to various control points, providing a smooth trajectory while ensuring that the fitted motion remained bounded in magnitude. This method allowed for twice-differentiable interpolation, facilitating realistic motion modeling [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. The fore flipper kinematics were validated using the robotic system to match the motions observed by the animal. This decomposition was not intended to challenge existing descriptions of the fore flipper stroke, but rather to supplement prior findings with a formulation that could be directly implemented on the robotic platform for the present experiments.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe bending of the fore flippers was identified as an important factor for propulsion and was therefore analyzed to ensure it could be accurately modeled in the robotic system. Based on the work of Friedman and Leftwich [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e], along with additional tracking conducted for this study, it was found that most of the bending occurs at the wrist joint, with continuous flexion extending through the digits (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC), (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003e). The amount of flipper deformation will influence both the direction and magnitude of the propulsive forces they generate. As a result, accurately modeling fore flipper bending will be critical to the development of the bio-robotic system.\u003c/p\u003e\u003cp\u003eThe same two-dimensional tracking technique was used to capture the motions of the head, pelvis, and hind flippers during a banking turn. Head motion was tracked from the body to the tip of the nose, while pelvic motion was measured from the body to the base of the tail. During the maneuver, the head was observed to roll a maximum of approximately 90\u0026deg;, yaw 40\u0026deg;, and pitch 90\u0026deg;, while the pelvis rolled 20\u0026deg;, yawed 30\u0026deg;, and pitched 80\u0026deg;. Throughout the turn, the hind flippers yawed 35\u0026deg; and rolled 60\u0026deg; [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. It remained unclear whether the pitching of the hind flippers resulted from joint articulation or passive flexibility of the flippers. Together, these measurements offered important guidance for modeling the sea lion\u0026rsquo;s turning mechanics in a bio-robotic system.\u003c/p\u003e"},{"header":"4 Bio-Robotic System Design","content":"\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e4.1 Overview\u003c/h2\u003e\u003cp\u003eBased on the observations and analyses discussed in Section 2 and 3, the Stroke Experimentation and Maneuver Optimizing Underwater Robot (SEAMOUR) was developed as a versatile research platform. Its design enables investigations into various swimming and maneuvering strategies through the actuation of multiple control surfaces, including the head, fore flippers, pelvis, and hind flippers (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). SEAMOUR was designed with features to proportionally match those of a sea lion, with a total length of 1.3m. The system features 14 actively controlled degrees of freedom (DOF), powered by high-torque waterproof servo motors (WR-7701, Xpert RC, USA). These include\u0026thinsp;\u0026plusmn;\u0026thinsp;60\u0026deg; of pitch and yaw at both the head and pelvis joints, 160\u0026deg; of roll\u003csub\u003ef\u003c/sub\u003e, 110\u0026deg; of yaw\u003csub\u003ef\u003c/sub\u003e, and 130\u0026deg; of pitch\u003csub\u003ef\u003c/sub\u003e at the fore flippers, and \u0026plusmn;\u0026thinsp;30\u0026deg; of yaw with 160\u0026deg; of roll at the hind flippers. This arrangement was selected to enable precise, repeatable, and independently controllable motions, providing a highly adjustable platform capable of delivering consistent performance across a broad range of swimming and maneuvering experiments.\u003c/p\u003e\u003cp\u003eSEAMOUR was developed as a modular research platform capable of producing biologically relevant motions while supporting a wide range of experimental configurations. To model sea lion body bending the head joint provides independent pitch and yaw actuation for modeling cervical bending, while the pelvic section uses the same mechanism to simulate lumbar motion. Each degree of freedom features independently adjustable actuation speeds and customizable motion profiles, allowing precise control of joint kinematics. The system\u0026rsquo;s modular design accommodates interchangeable fore flippers with adjustable shape and flexibility, along with interchangeable head shapes and tunable inertial properties. This adaptability allows SEAMOUR to evolve with new hypotheses and experimental needs without requiring a complete system redesign. The platform was engineered for reliable operation in both constrained and unconstrained testing environments, providing repeatable, consistent motions for controlled experimentation. Additionally, onboard sensing records the system rotations, enabling analysis of how the robot\u0026rsquo;s kinematics influence its body\u0026rsquo;s movement through the water.\u003c/p\u003e\u003cp\u003eCertain design simplifications were made to prioritize the features being investigated in these initial studies. Continuous body bending in the thoracic section, biologically accurate head shape, and exact flipper scaling and contour were excluded from this iteration of the design to focus on isolating the effects of the included features. Although these other features likely influence swimming performance, their omission allowed for a more controlled evaluation of the targeted kinematic variables. Future work will expand both numerical and experimental platforms to examine the influence of additional joints, body flexibility, and alternative control surface geometries on swimming and maneuvering performance.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e4.2 Main Body\u003c/h2\u003e\u003cp\u003eThe main body of the robotic platform, representing the thoracic region of the animal, was designed to enclose all mechanical and computational components within a streamlined ellipsoid shape, closely resembling that of a sea lion [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Unlike the animal, this section was made rigid in the current iteration of the system and does not allow for continuous body bending. The outer shell of the body was composed of multiple vacuum-formed sections made from 1.5 mm high-impact polystyrene (HIPS), which magnetically attach to internal support ribs for easy access to the underlying hardware. The front and rear shell sections taper smoothly from the circular head and pelvic regions to the oval-shaped central section. An aluminum U-channel serves as the structural core, providing rigidity and mounting points for internal components. SEAMOUR\u0026rsquo;s main body is divided into two sections: the anterior section, which provides mounting for the head and fore flippers, and the posterior section, which contains a waterproof box housing the electronics and provides a mounting point for the pelvic section.\u003c/p\u003e\u003cp\u003eA Raspberry Pi 4B (RPI) provided computational power and broad hardware compatibility, supporting motor control, sensor integration, and remote operation. A PWM-based daughter board (PN 2327, Adafruit, USA) was used for precise servo motor control, while a 9-DOF inertial measurement unit (IMU) (BNO085, PN 4754, Adafruit, USA) captured rotational and acceleration rates (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e). To enable remote operation, the RPI\u0026rsquo;s Wi-Fi antenna was modified so that it could be extended via a cable that breached the water\u0026rsquo;s surface. This antenna could either be dragged behind the robot using a surface float for continuous communication or remain attached to the outer shell after motor trajectories were loaded near the surface. The system was also designed with flexibility for future sensor integration, such as a depth sensor for active buoyancy control, a doppler velocity log for precise velocity and position measurements, or pressure sensors for local flow detection.\u003c/p\u003e\u003cp\u003ePower for the system was supplied by a battery and distributed through two voltage regulators: one dedicated to the RPI and another for the motors. A 2250 mAh 3-cell LiPo battery (MaxAmps, USA) provided a minimum of two hours of continuous operation. The battery was potted in epoxy resin and fitted with waterproof connectors on both the power and balance leads, allowing it to be slid into the anterior section of the U-channel core. Power from the battery supplied a 5V 3.2A regulator (D36V28F5, Pololu, USA) for the RPI and a 15A regulator (D24V150F9, Pololu, USA) for the 14 servos (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e). The servo power regulator was modified to allow it to be powered on and off via commands from the RPI, while the motor regulator was modified to operate at 8.4V, enabling each motor to reach its maximum performance of 8.2 Nm of torque and rotational speeds of 428 degrees per second.\u003c/p\u003e\u003cp\u003eAs the interior of the robot is flooded to enable the modular design, all electrical components were housed within a waterproof box located in the posterior section of the robot. This box was fabricated using a FDM 3D printer and coated with multiple layers of polyurethane to ensure waterproofing. The top opening was sealed with a 6 mm thick laser-cut acrylic sheet and a continuous 6 mm square neoprene gasket, fitted into a shallow groove to create a knife-edge seal. The lid was fastened with 4 mm bolts that connected to heat-set threaded inserts embedded from opposite side to prevent pullout. Antenna and servo motor wires were epoxy-potted into the box\u0026rsquo;s sides, securely sealing all penetrations (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e). This design allowed SEAMOUR to operate at depths up to 4m.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe system was designed so tests could be run while swimming freely or while attached to a float. Unconstrained tests were conducted utilizing an underwater docking station. This dock positioned the robot at a depth of 0.8 meters and featured two curved supports that matched the contour of the main body\u0026rsquo;s shell. Magnets embedded in a release rod connected to a steel bar running along the length of the posterior shell which allowed the system to be released remotely from the surface. This approach was used during both swimming and maneuvering tests conducted in this paper, which ensured a consistent starting position and orientation for each trial (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e). For constrained testing, the system was designed to be attached to a surface float, restricting its motion to the xy plane while preserving x axis rotations. To accomplish this a \u0026Oslash;160mm FDM 3D-printed Delrin roller bearing was fabricated and positioned between the front taper and main body shells. By attaching the system to a float, its position and orientation can be tracked at the surface and a surface docking system can be implemented for repeatable trials. This approach was not used in the studies presented in this paper, but its utility will be important for future studies.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e4.3 Fore Flippers\u003c/h2\u003e\u003cp\u003eThe fore flippers were designed to model the fundamental motions and bending of the biological sea lion flippers during characteristic swimming and turning maneuvers. Their motion was modeled at the elbow joint of the animal and actuated using three servo motors per flipper, providing independent control of pitch\u003csub\u003ef\u003c/sub\u003e, yaw\u003csub\u003ef\u003c/sub\u003e, and roll\u003csub\u003ef\u003c/sub\u003e. Described in the fore flipper frame of reference, the pitch\u003csub\u003ef\u003c/sub\u003e motor attaches to the center support structure, next in the chain is the yaw\u003csub\u003ef\u003c/sub\u003e motor, and lastly the roll\u003csub\u003ef\u003c/sub\u003e motor which acts as the final connection point for the fore flipper (Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003e). To increase the rigidity of the assembly, bearings were added on axis to the pitch\u003csub\u003ef\u003c/sub\u003e and roll\u003csub\u003ef\u003c/sub\u003e motors to support the radial loads. This assembly accommodates flippers of varying sizes and degrees of flexibility, while the available DOF enables a wide range of motions, ensuring the system is not limited to strictly biologically derived movements. For example, rigid flippers were used in place of the standard flexible flippers during the experimental trials conducted to validate the numerical model of the system, as discussed in Section 4.8 [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e].\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eSEAMOUR\u0026rsquo;s fore flippers were designed to model the general shape, proportions, and placement on the body. Certain biological characteristics, such as the trailing edge crenelations [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] and the specific hydrofoil shape [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] were simplified in the robotic system as these features were not the focus of these studies. The fore flippers were positioned to reflect the natural anatomy of a sea lion, but their size was reduced by 17% relative to a proportional scale to accommodate motor torque and speed limitations while maintaining key functional trends associated with propulsion and maneuvering. Preliminary tests showed that the hydrodynamic lift generated by the modeled fore flippers where similar to the actual sea lion fore flipper over a range of angles of attacks (unpublished data).\u003c/p\u003e\u003cp\u003eThe bending of the fore flippers was modeled based on the bending profile observed during the characteristic stroke. Each flipper was designed with a rigid base and a gridded support structure that enabled both span-wise and chord-wise bending to be modified during the design process (Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e13\u003c/span\u003e). The proportions of the rigid and flexible sections were derived from the animal's anatomy, where the forearm represented two-fifths and the hand portion represented three-fifths of the flipper (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB) [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. This design ensured the fabricated flippers modeled natural bending from the wrist joint through the finger-like structure [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Bending of the fore flipper was passively induced by self-loading during movement through water, with the degree of bending adjustable by modifying the taper of the grid pattern or the number of grid sections. Finite element analysis (FEA), conducted using SolidWorks CAD software (SolidWorks, USA), simulated the response of the flipper's support structure to a non-uniform distributed 5N load, modeling the flipper's motion through water (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003e). This simulation facilitated fine-tuning of the flipper\u0026rsquo;s structure before fabrication and validation on the mechanical system. The analysis predicted a tip displacement of 0.103 m, which when scaled closely matched the bending characteristics observed in the biological system (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe fore flippers were fabricated using a multi-step casting process (Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e13\u003c/span\u003e). First, a positive mold was created using an FDM 3D printer (F120, Stratasys, Minnesota USA). This mold was then vacuum-formed (450DT, Formech, USA) with clear 1.5 mm PETG sheets to produce the two halves of a negative mold. The 3D-printed support structures were positioned using a square mounting rod, which later served as the connection point to the motor assembly. The clear molds were fitted with a neoprene rubber gasket and secured using a 3D-printed framing clamp and bolts. A two-part silicone with a shore hardness of 30A (Dragon Skin 30, Smooth-On, USA) was poured into a preformed funnel at the top of the mold. The transparent molds allowed for visual confirmation that no air pockets were present in the silicone and that the support structure was properly aligned.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e4.4 Pelvis\u003c/h2\u003e\u003cp\u003eThe pelvic section was designed to emulate the bending of the animal\u0026rsquo;s lumbar region using a two-axis gimbal mechanism driven by two motors. This setup provides independent control over pitch and yaw movements. The gimbal is enclosed within a half-dome structure, with its axes aligned to a shared center of curvature, allowing up to \u0026plusmn;\u0026thinsp;60\u0026deg; of articulation from the midline in each direction. As minimal pelvic roll rotation was observed by the animal, rotation about this axis was excluded from the joint design. The pelvic section also integrates the motors responsible for hind flipper actuation and voids throughout are used for ballast material.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e4.5 Head\u003c/h2\u003e\u003cp\u003eSEAMOUR\u0026rsquo;s head was mounted on a two-axis gimbal system, modeling the degrees of freedom of a sea lion\u0026rsquo;s cervical region. In these initial studies, the head was axisymmetric to reduce mechanical complexity, measuring 0.265 m in length with a cross-sectional area of 0.02 m\u0026sup2;. To minimize drag and mitigate potential flow disruptions caused by head articulation, a four-way stretch spandex fabric was secured between the head and the tapered body shell (Fig.\u0026nbsp;\u003cspan refid=\"Fig14\" class=\"InternalRef\"\u003e14\u003c/span\u003eA). The head mounting system was designed for interchangeable head shapes and sizes, allowing for future investigations into the effects of head asymmetry. For example, a smaller axisymmetric head is utilized in Section 6.6, demonstrating this modularity. Beyond its adaptable form, the head also accommodates ballast material and space for a future vision system.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e4.6 Hind Flippers\u003c/h2\u003e\u003cp\u003eThe hind flippers of the bio-robotic system were designed to model the kinematics of the animal using flexible flippers that could yaw and roll independently. Housed within the pelvic section, the hind flipper motors enable each flipper to yaw a maximum of \u0026plusmn;\u0026thinsp;30\u0026deg; from midline and roll 160\u0026deg; (Fig.\u0026nbsp;\u003cspan refid=\"Fig14\" class=\"InternalRef\"\u003e14\u003c/span\u003eB). The flexible flippers were fabricated in the same manner as the fore flippers; however, their bending characteristics were not explicitly characterized, as the influence of their flexibility was not explored during these studies. When comparing the modeled hind flippers to the actual animal the hydrodynamic lift generated over a range of angles of attacks were similar (unpublished data).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e4.7 Trim\u003c/h2\u003e\u003cp\u003eLead weights and extruded polystyrene foam were added to designated voids in the head, body, and pelvis to adjust the locations of the center of gravity (COG) and center of buoyancy (COB), achieving neutral buoyancy and a horizontal equilibrium in the water. These adjustments positioned the COB above the COG in both the x-z and y-z planes (Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e15\u003c/span\u003eA), creating moments that affected the system's roll and pitch stability [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. The head, body, and pelvis sections were individually trimmed for neutrally buoyancy before being assembled. Trim modifications were also used to shift the COG and COB of the overall system anteriorly or posteriorly, with efforts made to align their positions close to the fore flippers as seen in the biological model [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Each one of these adjustments were updated in the 3D CAD model of SEAMOUR (Creo Parametric 10.0, USA), which was then used to calculate the COG of the system. In the final configuration, the COG for the entire system was calculated to be -0.07 m in the x-direction from the coordinate frame (Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e15\u003c/span\u003eB), which places it 0.12 m in front of the system's overall center.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTo calculate the COB, the solid parts and any air pockets in the CAD model needed to be set to a uniform density. Making these adjustments would allow the COG calculator in the 3D CAD software to be used to find the centroid of the displaced volume of fluid (i.e. COB). This calculation showed that the COB was also located \u0026minus;\u0026thinsp;0.07 m in the x-direction from the coordinate frame.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e4.8 Numerical Model\u003c/h2\u003e\u003cp\u003eComplementary numerical models of SEAMOUR were developed to simulate, analyze and visualize the coordinated flipper and body motions of the robotic system in water [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e], [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. Utilizing both Euler-Poincar\u0026eacute; and Newton-Euler formulations, these models are used to investigate and evaluate the bio-robotic system's maneuverability and stability across different swimming modes. To simulate the fluid forces, drag, lateral, lift and added mass forces were incorporated into the numerical model. The hydrodynamic coefficients, essential for calculating these forces, were determined through computational fluid dynamics (CFD) simulations and analytical techniques such as strip theory. These models were validated against SEAMOUR for various swimming and maneuvering trials. These models allow for the examination of the system's maneuverability, exploration of new strategies for propulsion, and assessment of design modifications prior to implementation on SEAMOUR.\u003c/p\u003e\u003c/div\u003e"},{"header":"5 Experimental Procedure and Data Collection","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003e5.1 Overview\u003c/h2\u003e\u003cp\u003eExperiments were conducted to implement and evaluate various swimming and maneuvering strategies using the bio-robotic system. These tests were designed to characterize the system\u0026rsquo;s dynamics, assess the use of various control surfaces, and identify performance trends as the robot moved freely through the water. Prior to testing, the system was trimmed to be horizontally level, with individual body sections trimmed for neutral buoyancy. The first set of experiments examined how the systems current COG and COB influenced its ability to return to equilibrium after being displaced at various angles (N\u0026thinsp;=\u0026thinsp;120), with stability assessed by measuring the response and settling times as the body realigned. The second set of experiments evaluated how different combinations of fore flipper and body motions could be used to achieve rectilinear swimming (N\u0026thinsp;=\u0026thinsp;15), with body orientation and x- and z-axis translations recorded to assess each stroke\u0026rsquo;s ability to maintain straight and level trajectories. The third set of trials investigated the use of head and pelvis motions to perform controlled pitch and yaw maneuvers (N\u0026thinsp;=\u0026thinsp;60), which are assessed by recording body orientation to quantifying angular positions and rates as control surfaces were actuated through a range of angles. Finally, insights from these experiments were combined to develop coordinated swimming and maneuvering sequences, demonstrating the full versatility and modularity of the system.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e5.2 Passive Roll and Pitch Stability\u003c/h2\u003e\u003cp\u003eThe impact of the system's COG and COB locations on its roll and pitch stability were determined by observing the body's angular rate as it returned to equilibrium. Experiments were conducted by manually positioning the robot at various roll and pitch angles before releasing it. An onboard IMU recorded the resulting angular changes over time at a sampling rate of 20 Hz (Fig.\u0026nbsp;\u003cspan refid=\"Fig16\" class=\"InternalRef\"\u003e16\u003c/span\u003e). Tests were conducted with roll and pitch angles of 20\u0026deg;, 40\u0026deg;, 60\u0026deg;, and 80\u0026deg;. Pitch tests were performed with the system in a streamlined position, with the fore flippers held against the body. Roll tests were conducted under two conditions: with the fore flippers streamlined and with the fore flippers extended (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\mathbf{y}\\mathbf{a}\\mathbf{w}\\mathbf{e}\\mathbf{d}}_{\\varvec{f}}\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\mathbf{r}\\mathbf{o}\\mathbf{l}\\mathbf{l}\\mathbf{e}\\mathbf{d}}_{\\varvec{f}}\\:\\)\u003c/span\u003e\u003c/span\u003e90\u0026deg;). The robot was tested by rolling it to the left and right, as well as pitching it in head-up and pelvic-up orientations. The data was offset to align t\u0026thinsp;=\u0026thinsp;0 with the release time and trimmed to 16 seconds. The systems pitch or roll orientation from three trials were averaged for analysis and the standard error between trials was calculated.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eSettling time was identified as the first instance when the signal entered and remained within a defined threshold band around the final value. This final value was estimated by fitting a linear trend line to the last 7 seconds of the roll data and extrapolating to the endpoint. The threshold band was set as \u0026plusmn;\u0026thinsp;0.1\u0026deg; from the final value, allowing for small fluctuations while capturing the signal's convergence behavior. The algorithm checked for continuous periods within this band that lasted for a minimum duration of 0.10 seconds to confirm stability. The first point at which the data satisfied these criteria was recorded as the settling time.\u003c/p\u003e\u003cp\u003eDue to the differences in the data set, a different methodology for calculating the settling time for pitch tests was needed. First the average of the final 10 data points of each test was calculated to represent the steady-state value. A\u0026thinsp;\u0026plusmn;\u0026thinsp;0.1\u0026deg; band was established around this steady-state value to define the acceptable range for settling. The settling time was identified as the first time point where the data enters this band and remains within it for the rest of the duration. This approach ensures that the system is considered settled only when it consistently stays close to its final value. These two approaches for determining settling time were necessary because SEAMOUR could not be precisely trimmed to achieve equilibrium at 0\u0026deg; in both the x and y axes.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e5.3 Swimming\u003c/h2\u003e\u003cp\u003eTo evaluate the impact of flipper and body motions on SEAMOUR's swimming capabilities, studies were conducted to assess various approaches for achieving rectilinear swimming. Three swimming techniques were tested for their effectiveness in producing straight and level locomotion: a biologically derived characteristic stroke (implemented both with and without pelvic pitching) and a modified side paddle stroke (Fig.\u0026nbsp;\u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e18\u003c/span\u003e). Developing a stroke that minimizes pitching and vertical displacement while maximizing forward velocity is essential for maintaining a steady trajectory and executing maneuvers at speed.\u003c/p\u003e\u003cp\u003eThe kinematics for the characteristic stroke were derived from existing literature, in-house kinematic investigations, and preliminary studies. While a separate future publication will detail a study investigating the lift and thrust forces produced during different phases of the characteristic sea lion stroke, the findings from that work provided kinematics that minimized lift and produced optimal thrust for these experiments. Additional trials utilized the pelvic section as a control surface to correct for pitching during the characteristic stroke. Preliminary tests determined that pitching the pelvic section downward by 15\u0026deg; during the stroke transition (between the power and paddle phases) and gradually returning it to a streamlined position during the recovery phase produced effective results.\u003c/p\u003e\u003cp\u003eThe modified side paddle stroke was designed to utilize only the fore flippers for rectilinear swimming. By manually adjusting the PCHIP spline control points for fore flipper motion, a new swimming gait was developed. This gait was adapted from the best-performing reinforcement-learned gait evaluated in earlier work [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e], with slight tuning to accommodate design modifications. This modified stroke replaces the power phase with a side-body paddle motion, allowing control of vertical body movement through flipper roll and pitch adjustments.\u003c/p\u003e\u003cp\u003eDuring each experiment, the position of SEAMOUR was tracked using a stationary underwater camera while orientation was recorded using an on board IMU. Each stroke type had a three second period with each trail consisting of four consecutive strokes. A GoPro Hero 11 camera (GoPro, USA), positioned to capture the side of the robot as it moved through the water, was set 8 m away and to record at 24 frames per second in 4K resolution using a wide frame of view. The video footage was then post-processed to correct lens distortion, lighting, and color (Adobe Premiere Pro, USA) before being tracked in Kinovea. To calibrate the scale of the video frame, a known length on SEAMOUR was set using the software. Pitch orientation data was collected during each trial using the IMU, which recorded at 20 Hz starting two seconds before and continuing throughout the trial; with the initial reading used to correct for any offset in the robot\u0026rsquo;s starting orientation. The position and orientation data from each study were aligned and results averaged across three trials, with the variability between them recorded as standard error.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003e5.4 Yaw and Pitch Turns\u003c/h2\u003e\u003cp\u003eTo evaluate the effectiveness of pitch and yaw turns on SEAMOUR, various combinations of the head and pelvis angles were explored. During these tests, SEAMOUR swam unconstrained, using its fore flippers to reach an initial velocity of 0.31 m/s at the start of each trial. Following this, the head, pelvis, or both were actuated to either 30\u0026deg; or 60\u0026deg; in yaw or pitch. During the combined tests, the head and pelvic were actuated symmetrically, i.e. head at 30\u0026deg; pitch and pelvic at 30\u0026deg; pitch. Pitch and yaw tests were performed about their respective axes\u0026mdash;pitch about the y-axis and yaw about the z-axis. Each pose was held for 5 seconds, and the IMU recorded the system\u0026rsquo;s orientation. A representative data set of three trails was then averaged, with the standard error between trials calculated. During each pose, the fore flippers remained in a streamlined position alongside the body, and the hind flippers were held at a roll angle of 0\u0026deg;.\u003c/p\u003e\u003c/div\u003e"},{"header":"6 Results and Discussion","content":"\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\u003ch2\u003e6.1 Overview\u003c/h2\u003e\u003cp\u003eSEAMOUR was successfully developed as a research platform to investigate the use of various biological and non-biological swimming and maneuvering gaits. Across all experimental trials, the system demonstrated consistent, repeatable performance with low positional and angular standard errors, validating SEAMOUR\u0026rsquo;s reliability and versatility as a modular, multi-bodied robotic platform. Initial static stability tests established a baseline, revealing that roll response varied with fore flipper configuration, as extended flippers improved stability by reducing angular velocity, settling time, and overshoot, while pitch maneuvers exhibited lower angular velocities but comparable settling times to roll tests with streamlined flippers. When implementing a biologically derived swimming stroke, known as the characteristic stroke, the system saw significant translation in heave and high pitch changes through the trials. When the pelvis was incorporated during the characteristic stroke, heave translation was mitigated, due to reduced pitching, and surge translation was greatly improved. Further refinement using a modified side paddle stroke eliminated the need for pelvic actuation while still achieving effective forward translation and maintaining a low body pitch angle. In maneuvering trials, the pelvic section had the greatest influence during pitch and yaw maneuvers, while combining head and pelvic articulation saw slight improvements. Overall, pitch-based maneuvers consistently outperformed yaw-based maneuvers under all test conditions. These findings collectively highlight the ability to implement various swimming and maneuvering gaits, which would pave the way for additional gait strategies tailored to specific locomotory objectives.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\u003ch2\u003e6.2 Influence of COG and COB\u003c/h2\u003e\u003cp\u003eAs SEAMOUR is a multi-bodied robot, the buoyancy characteristics of its head and pelvic sections can influence the performance of the overall system when articulated. If these independent sections were negatively buoyant, while the overall system was neutrally buoyant, any displacement of these sections from the body\u0026rsquo;s midline would have caused the COG to shift more than the COB, causing the system to roll. To mitigate this action, the head and pelvis sections were individually made neutrally buoyant to minimize the influence of hydrostatic forces, particularly during the maneuverability studies. This configuration ensures that the COG and COB approximately shift in unison when individual body sections are articulated.\u003c/p\u003e\u003cp\u003eThe COG, COB, and moments of inertia are closely interconnected, and their combined configuration significantly affects the maneuverability of the bio-robotic platform. In a configuration where the head and pelvic sections are negatively buoyant while the overall system remains neutrally buoyant, the moments of inertia about the y and z axes will be greater compared to a configuration where each section was independently neutrally buoyant. The former configuration resembled a dumbbell, while the latter was akin to a pipe. Consequently, a negatively buoyant head and pelvis increase the system\u0026rsquo;s pitch and yaw stability, making it suitable for applications prioritizing stable straight swimming. However, this setup is less optimal for a maneuverable system, further justifying the trimming of the COG/COB to enhance maneuverability. Understanding and optimizing the moments of inertia are essential in the design of effective underwater vehicles, as the offset and relative positions of the COG and COB directly influence a system\u0026rsquo;s stability and the ability to roll, yaw, and pitch. These parameters can be strategically adjusted to achieve specific performance characteristics depending on the operational goals of the vehicle. Future work will investigate the effects of modifying these properties through additional testing.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\u003ch2\u003e6.3 Passive Roll and Pitch Stability\u003c/h2\u003e\u003cp\u003eDuring tests when the system was rolled away from equilibrium, its response varied significantly based on the fore flipper configuration but demonstrated consistent trends across different initial roll angles (Fig.\u0026nbsp;\u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e18\u003c/span\u003e). Larger initial roll angles resulted in greater overshoot magnitudes when the fore flippers were streamlined, whereas the opposite occurred when the fore flippers were extended. By extending the fore flippers, the angular velocity during the first second is decreased by 66% across all four angles (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The mean standard error across both roll configurations ranged from \u0026plusmn;\u0026thinsp;0.17\u0026deg; to \u0026plusmn;\u0026thinsp;0.47\u0026deg; across each 16 second trail (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), demonstrating consistent performance.\u003c/p\u003e\u003cp\u003eWith the fore flippers in the streamlined configuration, the roll angle underwent multiple cycles of damped oscillations before settling, while in the extended configuration, stability was achieved after a single damped oscillation. The settling time in Fig.\u0026nbsp;\u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e18\u003c/span\u003eA and B is denoted by a black dot on each line. This variation in the system's response can be attributed to the difference in drag between configurations. Notably, when comparing the two roll configurations, the settling time was significantly reduced when the fore flippers were extended (abducted). Comparing the streamlined and extended positions revealed that extending the fore flippers reduced settling time by 53% for 40\u0026deg;, 53% for 60\u0026deg;, and 49% for 80\u0026deg; initial roll angles.\u003c/p\u003e\u003cp\u003eWith the fore flippers extended during the roll test the settling time for release angles of 40\u0026deg;, 60\u0026deg;, and 80\u0026deg; consistently fell between six and seven seconds, while the 20\u0026deg; release angle deviated from this pattern. This was likely due to the system's low restoring moment at smaller initial angles, causing it to settle slightly off the target position after the first overshoot. Consequently, the extrapolated trend line predicted a final value that was farther from the expected position at 15.3s.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eMean angular velocity during the first second of each trial for fore flipper streamlined and extended configurations at 20\u0026deg;, 40\u0026deg;, 60\u0026deg;, and 80\u0026deg; roll angles from equilibrium.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"7\" nameend=\"c7\" namest=\"c1\"\u003e\u003cp\u003eMean angular velocity\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eFore Flipper Configuration\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e20\u0026deg; \u003cem\u003e(\u0026deg;/s)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e40\u0026deg; \u003cem\u003e(\u0026deg;/s)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e\u003cp\u003e60\u0026deg; \u003cem\u003e(\u0026deg;/s)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e80\u0026deg; \u003cem\u003e(\u0026deg;/s)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStreamline\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e30\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e49\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e61\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e70\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eExtended\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e24\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eMean standard error for fore flipper streamlined and extended configurations at 20\u0026deg;, 40\u0026deg;, 60\u0026deg;, and 80\u0026deg; roll from equilibrium.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eMean standard error\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eFore Flipper Configuration\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e20\u0026deg; \u003cem\u003e(\u003c/em\u003e\u0026plusmn;\u003cem\u003e\u0026deg;)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e40\u0026deg; \u003cem\u003e(\u003c/em\u003e\u0026plusmn;\u003cem\u003e\u0026deg;)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e60\u0026deg; \u003cem\u003e(\u003c/em\u003e\u0026plusmn;\u003cem\u003e\u0026deg;)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e80\u0026deg; \u003cem\u003e(\u003c/em\u003e\u0026plusmn;\u003cem\u003e\u0026deg;)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStreamline\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.47\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eExtended\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.33\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.42\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.35\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eWhen the robot was pitched away from its equilibrium and released, it had lower mean angular velocities compared to both roll tests (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e), but had comparable settling times to the roll configuration with the flippers in the streamlined position (Fig.\u0026nbsp;\u003cspan refid=\"Fig18\" class=\"InternalRef\"\u003e19\u003c/span\u003e). The increase in drag and moment of inertia during pitch tests resulted in lower angular velocities, preventing the system from overshooting its final value in this orientation. Over the 16 second trials the mean standard error ranged from \u0026plusmn;\u0026thinsp;0.66\u0026deg; to \u0026plusmn;\u0026thinsp;1.26\u0026deg; (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eMean angular velocity during the first second of each trial for fore flipper streamlined and extended configurations at 20\u0026deg;, 40\u0026deg;, 60\u0026deg;, and 80\u0026deg; pitch angles from equilibrium.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"7\" nameend=\"c7\" namest=\"c1\"\u003e\u003cp\u003eMean angular velocity\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eFore Flipper Configuration\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e20\u0026deg; \u003cem\u003e(\u0026deg;/s)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e40\u0026deg; \u003cem\u003e(\u0026deg;/s)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e\u003cp\u003e60\u0026deg; \u003cem\u003e(\u0026deg;/s)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e80\u0026deg; \u003cem\u003e(\u0026deg;/s)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStreamline\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.86\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.89\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003e2.68\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u003cp\u003e4.84\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eMean standard error for fore flipper streamlined and extended configurations at 20\u0026deg;, 40\u0026deg;, 60\u0026deg;, and 80\u0026deg; pitch from equilibrium.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003eMean standard error\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eFore Flipper Configuration\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e20\u0026deg; \u003cem\u003e(\u003c/em\u003e\u0026plusmn;\u003cem\u003e\u0026deg;)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e40\u0026deg; \u003cem\u003e(\u003c/em\u003e\u0026plusmn;\u003cem\u003e\u0026deg;)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e60\u0026deg; \u003cem\u003e(\u003c/em\u003e\u0026plusmn;\u003cem\u003e\u0026deg;)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e80\u0026deg; \u003cem\u003e(\u003c/em\u003e\u0026plusmn;\u003cem\u003e\u0026deg;)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStreamline\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.66\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.82\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.20\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe results demonstrated that SEAMOUR\u0026rsquo;s roll response resembled an underdamped system, while its pitch acted more like a critically damped system. This response was consistent with the smaller moment of inertia about the roll axis and the reduced drag encountered in that direction. As roll maneuvers will not be explored in these studies, a higher static roll stability will help isolate the effects of control surfaces on pitch and yaw maneuvers. Additionally, the low static pitch stability ensured that pitch maneuvers remain largely unaffected, allowing for a clearer evaluation of rectilinear swimming performance in terms of both thrust and lift without significant counteracting effects from static pitch stability.\u003c/p\u003e\u003cp\u003eThe evaluation of SEAMOUR's passive roll and pitch stability established a critical baseline for understanding its inherent static stability, which must be assessed before investigating the system\u0026rsquo;s maneuvering abilities, particularly when modifications are made to its COG and COB. This baseline served as a reference for assessing how static stability influenced the system\u0026rsquo;s performance during dynamic tests. Specifically, these investigations help determine whether stability enhances or constrains the effectiveness of the control surfaces. This distinction is key to the subsequent discussion on the system\u0026rsquo;s maneuvering capabilities and the role of static stability in shaping overall performance.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e\u003ch2\u003e6.4 Swimming\u003c/h2\u003e\u003cp\u003eWhen a biologically derived characteristic stroke was implemented on SEAMOUR, pitching and heave translation affected its other all surge translation. Over the course of four strokes, the system pitched up by a total of 36\u0026deg;, with an average of 23\u0026deg;. This stroke also saw heave translation of 0.38m (\u0026plusmn;\u0026thinsp;0.03 m) and surge translation of 2.89 m (\u0026plusmn;\u0026thinsp;0.08m) (Fig.\u0026nbsp;\u003cspan refid=\"Fig19\" class=\"InternalRef\"\u003e20\u003c/span\u003e). The observed pitching resulted from the lift forces generated in combination with the COG location, due to a moment arm. This shows that pitch stability can be particularly critical during propulsive strokes. Since SEAMOUR\u0026rsquo;s static pitch stability is low, this stroke easily overpowered it. As the system continued pitching upward, increased surface area was exposed to the forward flow, further exacerbating pitching in subsequent strokes.\u003c/p\u003e\u003cp\u003eThe performance of the biologically derived stroke on SEAMOUR highlighted both the stroke\u0026rsquo;s limitations in straight swimming and its potential for maneuverability. Due to key differences between SEAMOUR and a biological animal, the stroke did not achieve effective straight swimming. However, it demonstrated how fore flipper lift forces generate pitching moments, which may be useful for maneuvers. Fortunately, the fore flippers are not restricted to a single set of kinematics, allowing their trajectories to be modified or additional control surfaces to be incorporated during the stroke to achieve straight swimming (Fig.\u0026nbsp;\u003cspan refid=\"Fig20\" class=\"InternalRef\"\u003e21\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eTo minimize pitching and heave translation during the characteristic stroke, the pelvic section was used during the stroke, resulting in better surge translation and less overall pitching. This stroke maintained an average pitch angle of 9\u0026deg; and reduced the final pitch angle by 63% after four strokes, compared to trials without pelvic actuation. The system achieved a final horizontal translation of 3.24m (\u0026plusmn;\u0026thinsp;0.09 m) and a vertical translation of -0.05m (\u0026plusmn;\u0026thinsp;0.04m). These results highlight how the coordinated use of additional control surfaces during a fore flipper propulsive stroke can balance forces and produce straight swimming.\u003c/p\u003e\u003cp\u003eThe modified side paddle stroke had the least amount of pitching and comparable surge translation as the characteristics stroke with the pelvic actuation. This stroke saw horizontal translation of 3.23 m (\u0026plusmn;\u0026thinsp;0.08 m) while maintaining an average pitch angle of 2\u0026deg; over four strokes. An increase in pitch angle between seven and twelve seconds was observed, potentially due to drag from the surface antenna. Slight negative buoyancy was evident, as indicated by minimal pitching during the first three strokes while maintaining a consistent downward trend in the Z direction, resulting in a final vertical translation of -0.16 m (\u0026plusmn;\u0026thinsp;0.06 m). This approach demonstrates how modifying the 3D kinematics of the fore flipper can achieve effective straight swimming, and by eliminating the need for additional control surfaces during fore flipper propulsion, these surfaces can be reallocated for coordinated maneuvers, as discussed in Section 6.6.\u003c/p\u003e\u003cp\u003eLow standard errors were observed in position and orientation data during various swimming and maneuvering trials. As the system swam unconstrained for 12 seconds, covering 3 meters of horizontal translation, the standard error for both horizontal and vertical displacement remained below 8% of SEAMOUR\u0026rsquo;s total length. During these trials the standard error for the pitch angle was 1.5\u0026deg; for the characteristic stroke, 0.8\u0026deg; for the characteristic stroke with pelvic actuation, and 2.2\u0026deg; for the modified paddle stroke. These low error margins demonstrate the system\u0026rsquo;s ability to have repeatable swimming performance across different gaits and configurations.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec22\" class=\"Section2\"\u003e\u003ch2\u003e6.5 Pitch and Yaw Maneuvers\u003c/h2\u003e\u003cp\u003eThis section evaluates the system's trends and performance impacts. While a comprehensive quantitative analysis is reserved for a future publication, the results presented here highlight key trends in the system's performance. Quantitative data, specifically standard errors, are used to demonstrate the system's repeatability and consistency across different maneuvers and test conditions.\u003c/p\u003e\u003cp\u003eDuring pitch maneuvers, simultaneously actuating the head and pelvis at higher angles yielded greater mean rotational rates and final angular positions, with the pelvis contributing more significantly than the head (Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Comparing final angular positions, the pelvic section consistently outperformed the head section at both tested pose angles. Increasing the actuation angle from 30\u0026deg; to 60\u0026deg; produced a similar increase in rotation for the pelvis and combined sections, while the head section exhibited a more modest improvement. When the head and pelvic sections were actuated together, the resulting angular positions closely matched those of the pelvic section alone, with only minor additional increases at both pose angles.\u003c/p\u003e\u003cp\u003eFor yaw maneuvers, the pelvic section consistently outperformed the head section when actuated independently, while the combination of both produced the highest final angular positions. The pelvic section achieved substantially greater rotations than the head at both tested pose angles, and when actuated together, the head and pelvic sections further improved performance beyond the pelvic section alone, with more noticeable gains at the lower pose angle. Increasing the actuation angle from 30\u0026deg; to 60\u0026deg; produced varied outcomes: the head section showed minimal improvement, while both the pelvic and combined sections demonstrated substantial increases in final angular positions. In these yaw studies, the head contributed more effectively when paired with the pelvic section. This could be attributed to the hind flippers' position during the maneuver. Since they were rolled flat in these tests, they cut through the water, reducing the pelvic section's effectiveness and allowing the head's motion to exert a relatively greater influence on the generated forces.\u003c/p\u003e\u003cp\u003eWhen comparing pitch and yaw maneuvers, pitch consistently outperformed yaw in both final angular position and average turning rates (Fig.\u0026nbsp;\u003cspan refid=\"Fig21\" class=\"InternalRef\"\u003e22\u003c/span\u003e) (Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Head articulation at both pose angles resulted in relatively low final angular positions for both pitch and yaw. In contrast, pelvic articulation during pitch produced noticeably higher final positions than yaw at both tested angles. When the head and pelvis were actuated together, pitch maneuvers demonstrated greater increases in final angular position compared to yaw, with larger gains observed at the higher pose angle. Average turning rates for head articulation were similar between pitch and yaw. However, pelvic articulation favored pitch, with notably higher turning rates at both pose angles. When combining the head and pelvis, pitch maneuvers again produced substantially higher turning rates compared to yaw, with the difference more pronounced at the lower pose angle. Detailed values for mean turning rates are presented in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e.\u003c/p\u003e\u003cp\u003eClear trends were evident across both pitch and yaw studies, depending on which sections were actuated. During all maneuvers, the pelvic section consistently outperformed the head, likely due to the axisymmetric shape of the head and the inclusion of large, flat hind flippers on the pelvic section, which additionally enhance its effectiveness during pitch maneuvers. For both 30\u0026deg; and 60\u0026deg; maneuvers, the pelvic section alone initially outperformed the combined actuation of the head and pelvis. In the first 0.5 seconds, the pelvic section achieved a larger initial angular change compared to the more gradual response of the combined pose. This trend was observed in both pitch and yaw studies, with a more pronounced effect during yaw tests. This pattern becomes clearer when considering the results of head actuation alone, which consistently produced an initial negative angular position in both pitch and yaw maneuvers before transitioning to a positive angle. During head actuation, the motor, located anterior to the system's COG, imparts a reaction moment. Consequently, the main body, containing the IMU, experiences an initial negative angular displacement. This effect can be seen most dramatically in the head pitch and yaw 60\u0026deg; trials. In contrast, pelvic section actuation induces the opposite effect, which can be seen most prominently in the pelvic 60\u0026deg; yaw plot. Unlike the yaw studies, which show a relatively linear increase, the pitch studies are influenced by restorative forces resulting from the offset between the COG and COB.\u003c/p\u003e\u003cp\u003eThe findings revealed low standard errors across trials and configurations, demonstrating consistent trends. Since standard error propagated over the length of the trials, examining the error at the final 3 seconds provided insight into overall performance. At 3 seconds, the final pitch error was \u0026plusmn;\u0026thinsp;2.0\u0026deg; and \u0026plusmn;\u0026thinsp;0.9\u0026deg; for head-alone trials at 30\u0026deg; and 60\u0026deg;, respectively, \u0026plusmn;\u0026thinsp;0.4\u0026deg; and \u0026plusmn;\u0026thinsp;0.3\u0026deg; for pelvis-alone trials, and \u0026plusmn;\u0026thinsp;0.3\u0026deg; and \u0026plusmn;\u0026thinsp;0.4\u0026deg; when the head and pelvis were used together. For yaw tests, the final error for head-alone trials was \u0026plusmn;\u0026thinsp;2.4\u0026deg; and \u0026plusmn;\u0026thinsp;2.0\u0026deg;, for pelvis-alone trials was \u0026plusmn;\u0026thinsp;2.5\u0026deg; and \u0026plusmn;\u0026thinsp;3.1\u0026deg;, and for combined actuation was \u0026plusmn;\u0026thinsp;3.7\u0026deg; and \u0026plusmn;\u0026thinsp;0.8\u0026deg; at actuation angles of 30\u0026deg; and 60\u0026deg;, respectively. These results indicate that the system maintained a low error range, varying from as little as \u0026plusmn;\u0026thinsp;0.3\u0026deg; to a maximum of \u0026plusmn;\u0026thinsp;3.7\u0026deg; over the three-second turning period as it maneuvered through 3D space, demonstrating the system\u0026rsquo;s repeatable performance.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eMean turning rate for pitch and yaw trials over the 3s trails\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e\u003cp\u003eMean Turning Rates\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003ePose\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePitch \u003cem\u003e(\u0026deg;/s)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYaw \u003cem\u003e(\u0026deg;/s)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHead 30\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3.17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4.12\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHead 60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5.23\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5.33\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePelvic 30\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e12.69\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e8.38\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePelvic 60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e22.99\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e13.45\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHead and Pelvic 30\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e11.88\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5.51\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHead and Pelvic 60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e21.77\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e11.67\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec23\" class=\"Section2\"\u003e\u003ch2\u003e6.6 Coordinated Maneuvers\u003c/h2\u003e\u003cp\u003eInsights gained from the experimental swimming and maneuvering trials provided a foundational understanding to produce coordinated maneuvers using flippers and various body segments. These coordinated maneuvers used various forms of flipper and body motions to achieve roll, yaw, and pitch maneuvers. These maneuvers followed pre-prescribed motor kinematics, meaning the system operated without feedback during its movements. For each test, SEAMOUR was positioned at a depth suitable for the intended trajectory.\u003c/p\u003e\u003cp\u003eSEAMOUR's versatile maneuvering strategy, involving coordinated movements of its head, pelvis, and fore flippers, allowed it to effectively follow a complex curvilinear path. Figure\u0026nbsp;\u003cspan refid=\"Fig23\" class=\"InternalRef\"\u003e24\u003c/span\u003e illustrates a 20-second sequence with the robot\u0026rsquo;s position recorded every 4 seconds and overlayed onto a single image. To execute this sequence, SEAMOUR first accelerated using two modified side paddle strokes. The head and pelvis were then pitched down 30\u0026deg; while another stroke to initiate a dive. With the system pitched downward, the head and pelvis returned to a streamlined position, and an additional stroke was used to drive SEAMOUR deeper at a 25\u0026deg; angle. Subsequently, the head and pelvis were pitched up 30\u0026deg; during an additional stroke to change direction upward. Following this, the head and pelvis returned to the streamlined position, and another stroke was performed to continue the ascent at a 40\u0026deg; angle. To level the body, the head and pelvis were pitched down 30\u0026deg; during a subsequent stroke, after which SEAMOUR completed additional modified side paddle strokes to swim out of the camera\u0026rsquo;s view. This sequence demonstrates the effective use of the head, pelvis, and fore flippers to perform a coordinated series of maneuvers so swim through a curvilinear path.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eBy leveraging the modularity of the system, adjustments to head size and the COG/COB distribution were made to reduce roll stability to execute the three coordinated maneuvers seen in Fig.\u0026nbsp;\u003cspan refid=\"Fig24\" class=\"InternalRef\"\u003e25\u003c/span\u003e. The head and pelvic sections were designed to be negatively buoyant while maintaining overall neutral buoyancy. This configuration caused the COG to shift laterally when these sections yawed, generating restorative forces that rolled the body to vertically realign the COG and COB. The z distance between the COG and COB was reduced while still retaining the system horizontal orientation in water. Additionally, a smaller axisymmetric head was added as less volume was required for these hydrostatic adjustments.\u003c/p\u003e\u003cp\u003eSEAMOUR demonstrated a 360\u0026deg; roll maneuver using only a single flipper, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig24\" class=\"InternalRef\"\u003e25\u003c/span\u003eA. To execute this, the robot first accelerated with three modified side paddle strokes. Immediately following, a single flipper executed three consecutive strokes. These strokes leveraged the fore flipper's roll angle during both its abduction and adduction phases, resulting in a rotation about the bodies roll axis. Beyond this, traditional asymmetric fore flipper roll angles have also been explored for body rolling, a technique used in the coordinated maneuvers shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig24\" class=\"InternalRef\"\u003e25\u003c/span\u003eB and Fig.\u0026nbsp;\u003cspan refid=\"Fig24\" class=\"InternalRef\"\u003e25\u003c/span\u003eC.\u003c/p\u003e\u003cp\u003eSEAMOUR executed a biologically derived banking turn (Fig.\u0026nbsp;\u003cspan refid=\"Fig24\" class=\"InternalRef\"\u003e25\u003c/span\u003eB), modeling the maneuver observed in the California sea lion (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The maneuver began with the fore flippers \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{y}\\text{a}\\text{w}\\text{e}\\text{d}}_{f}\\)\u003c/span\u003e\u003c/span\u003e 90\u0026deg; and asymmetrically \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{r}\\text{o}\\text{l}\\text{l}\\text{e}\\text{d}}_{f}\\)\u003c/span\u003e\u003c/span\u003e, creating an effective offset of 40\u0026deg;, where one flipper is angled with the top face into flow and the other with the bottom face in the direction of flow. Simultaneously the head and pelvic section were yawed 60\u0026deg; while the hind flippers were rolled flat and yawed outward by 20\u0026deg;. As the body rolls, the head and pelvic sections adjusted their yaw and pitch angle to maintain alignment with the xy plane, reaching 60\u0026deg; of head/ pelvic pitch when the body is rolled 90\u0026deg;. The fore flippers were then \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{r}\\text{o}\\text{l}\\text{l}\\text{e}\\text{d}}_{f}\\)\u003c/span\u003e\u003c/span\u003e to 90\u0026deg; during the remainder of the turn. To return the body to its initial state, the head and pelvic sections were streamlined as the lower fore flipper was flapped to roll the body, and the other fore and hind flippers slowly return to their streamlined position. This highlights SEAMOUR\u0026rsquo;s ability to execute maneuvers modeled after the sea lion, allowing for deeper investigation into biological approaches.\u003c/p\u003e\u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig24\" class=\"InternalRef\"\u003e25\u003c/span\u003eC, SEAMOUR executes a 180\u0026deg; pitch turn by performing a front flip type maneuver. This maneuver was achieved by first using the characteristic stroke with pelvic actuation to accelerate the system. Subsequently, the head and pelvic sections were pitched downward by 60\u0026deg; as five modified side paddle strokes were executed to flip the body 180\u0026deg;. With the fore flippers then extended, they were \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{r}\\text{o}\\text{l}\\text{l}\\text{e}\\text{d}}_{f}\\)\u003c/span\u003e\u003c/span\u003e asymmetrically, causing the body to roll 180\u0026deg; back to equilibrium before performing further rectilinear swimming strokes. These coordinated maneuvers highlight SEAMOUR's versatility and demonstrate the successful integration of key findings from these studies.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"7 Conclusion and Future Work","content":"\u003cp\u003eA bio-robotic underwater system, inspired by the California sea lion, was successfully developed and validated as a capable and reliable research platform. Through a series of comprehensive swimming and maneuvering experiments, this paper has demonstrated the system's foundational design and its effectiveness as a research tool. These tests showcased the platform's versatility and repeatability, supported by a modular design built for expanding hypotheses. The system uniquely facilitates both independent and combined evaluations of various factors contributing to its performance, as demonstrated through coordinated swimming and complex roll, yaw, and pitch maneuvers. Beyond exploring biologically derived swimming gaits, the platform can also be tuned to investigate diverse hand-tuned or reinforcement-learned gaits. Ultimately, SEAMOUR stands as a versatile and reliable platform specifically designed to advance the exploration of morphological and kinematic traits contributing to effective underwater swimming and maneuvering.\u003c/p\u003e\u003cp\u003eWhile the system achieved a maximum velocity of 0.31 m/s during the swimming trails, SEAMOUR\u0026rsquo;s modular design allows for modifications that can increase swimming speed. Several approaches can be explored, such as developing swimming gaits specifically tailored to the robotic platform, designing flippers with larger surface area, and replacing the current motors with models offering higher speed and torque capabilities. Additionally, incorporating supplementary control surfaces to assist with propulsion could further increase the system\u0026rsquo;s velocity. Future work will investigate modifications to optimize the platform\u0026rsquo;s swimming performance for a broader range of experimental applications.\u003c/p\u003e\u003cp\u003eTo gain a deeper understanding of the impact of various control surfaces on pitch and yaw maneuvers, a more detailed investigation is necessary. By constraining the system to so that yaw and pitch turns can be conducted on the same plane, eliminating any restoring moments during pitch tests, a better comparison between pitch and yaw turns can be facilitated. Expanding the range of actuation angles would also provide valuable insights into how these control surfaces influence maneuvering capabilities. Additional tests should also assess the contribution of the fore flippers on turning performance. To further evaluate the effectiveness of pitch and yaw maneuvers, position data should be collected, enabling the calculation of turning radii for each maneuver and providing a more comprehensive understanding of the system's maneuverability. Having a better understanding of the system\u0026rsquo;s maneuverability will provide valuable insights into the system's potential in dynamic flow environments. These evaluations can provide insights into how specific features of the animal's swimming and maneuvering can offer potential strategies to enhance the maneuverability of UUVs.\u003c/p\u003e\u003cp\u003eActive buoyancy control should be integrated into the system's design to enable tests that isolate experimental factors and yield more repeatable results. During the current testing, constant buoyancy changes presented challenges. This was primarily because many components, fabricated using FDM 3D printing, experienced gradual water saturation, leading to a shift toward negative buoyancy over time. To compensate, slight adjustments were manually made to the main body's buoyancy using foam throughout the studies. Future enhancements should incorporate an active buoyancy management system, which would either maintain consistent buoyancy during operation or enable lateral buoyancy control for pitching the system.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eAcknowledgments\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThis research was funded by the Office of Naval Research (Dr. Tom McKenna, Program Officer, ONR Code 341). The funder played no role in study design, data collection, analysis and interpretation of data, or the writing of this manuscript. We would like to thank the work conducted by students in the Liquid Life Lab at West Chester University, as well as the Biologically Inspired Energy Laboratory at George Washington University. The authors would like to thank Anthony Paul Bibeck and Ahmet Yalim Kiral for their contributions in conducting bio-robotic experiments and collecting valuable data.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eData Availability\u003c/p\u003e\n\u003cp\u003eThe raw datasets generated and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003eCode availability\u003c/p\u003e\n\u003cp\u003eThe underlying code for this study is not publicly available but may be made available to qualified researchers on reasonable request from the corresponding author.\u003c/p\u003e\n\u003cp\u003eAuthor Contributions\u003c/p\u003e\n\u003cp\u003eN.M designed and developed the robotic system. N.M., S.K., and A.D. conducted studies with the robotic system. F.E.F provided biological insights to the California sea lion. M.C.L. provided information on the fluid dynamic interaction of the fore flippers. H.G.K worked with S.K on the development of the numerical models of the robotic system. J.L.T advised the building and testing of the robotic system. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003eCompeting Interests\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003eAdditional information\u003c/p\u003e\n\u003cp\u003eCorrespondence and requests for materials should be addressed to Nicholas Marcouiller ([email protected]).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003e\u0026ldquo;ONR Technology and Research,\u0026rdquo; Office of Naval Research. Accessed: Feb. 25, 2025. [Online]. Available: https://p74l1103a01.dc3n.navy.mil/our-research/onr-technology-and-research\u003c/li\u003e\n\u003cli\u003e\u0026ldquo;New Blue Economy | National Oceanic and Atmospheric Administration.\u0026rdquo; Accessed: Feb. 25, 2025. [Online]. Available: https://www.noaa.gov/blue-economy\u003c/li\u003e\n\u003cli\u003eR. Thandiackal and G. V. Lauder, \u0026ldquo;How zebrafish turn: analysis of pressure force dynamics and mechanical work,\u0026rdquo; \u003cem\u003eJ. Exp. Biol.\u003c/em\u003e, vol. 223, no. 16, p. jeb223230, Aug. 2020, doi: 10.1242/jeb.223230.\u003c/li\u003e\n\u003cli\u003eF. E. Fish and J. T. Beneski, \u0026ldquo;Evolution and Bio-Inspired Design: Natural Limitations,\u0026rdquo; in \u003cem\u003eBiologically Inspired Design: Computational Methods and Tools\u003c/em\u003e, A. K. Goel, D. A. McAdams, and R. B. Stone, Eds., London: Springer, 2014, pp. 287\u0026ndash;312. doi: 10.1007/978-1-4471-5248-4_12.\u003c/li\u003e\n\u003cli\u003eP. R. Bandyopadhyay, \u0026ldquo;Trends in biorobotic autonomous undersea vehicles,\u0026rdquo; \u003cem\u003eIEEE J. Ocean. Eng.\u003c/em\u003e, vol. 30, no. 1, pp. 109\u0026ndash;139, Jan. 2005, doi: 10.1109/JOE.2005.843748.\u003c/li\u003e\n\u003cli\u003eF. E. Fish and G. V. Lauder, \u0026ldquo;Control surfaces of aquatic vertebrates: active and passive design and function,\u0026rdquo; \u003cem\u003eJ. Exp. Biol.\u003c/em\u003e, vol. 220, no. 23, pp. 4351\u0026ndash;4363, Dec. 2017, doi: 10.1242/jeb.149617.\u003c/li\u003e\n\u003cli\u003eG. Li, G. Liu, D. Leng, X. Fang, G. Li, and W. Wang, \u0026ldquo;Underwater undulating propulsion biomimetic robots: a review,\u0026rdquo; \u003cem\u003eBiomimetics\u003c/em\u003e, vol. 8, no. 3, Art. no. 3, Jul. 2023, doi: 10.3390/biomimetics8030318.\u003c/li\u003e\n\u003cli\u003eP. R. Bandyopadhyay, \u0026ldquo;Guest Editorial: Biology-Inspired Science and Technology for Autonomous Underwater Vehicles,\u0026rdquo; \u003cem\u003eIEEE J. Ocean. Eng.\u003c/em\u003e, vol. 29, no. 3, pp. 542\u0026ndash;546, Jul. 2004, doi: 10.1109/JOE.2004.833099.\u003c/li\u003e\n\u003cli\u003eR. Salazar, V. Fuentes, and A. Abdelkefi, \u0026ldquo;Classification of biological and bioinspired aquatic systems: A review,\u0026rdquo; \u003cem\u003eOcean Eng.\u003c/em\u003e, vol. 148, pp. 75\u0026ndash;114, Jan. 2018, doi: 10.1016/j.oceaneng.2017.11.012.\u003c/li\u003e\n\u003cli\u003eA. Wm. English, \u0026ldquo;Limb movements and locomotor function in the California sea lion (Zalophus californianus),\u0026rdquo; \u003cem\u003eJ. Zool.\u003c/em\u003e, vol. 178, no. 3, pp. 341\u0026ndash;364, 1976, doi: 10.1111/j.1469-7998.1976.tb02274.x.\u003c/li\u003e\n\u003cli\u003eS. D. Feldkamp, \u0026ldquo;Swimming in the California sea lion: morphometrics, drag and energetics,\u0026rdquo; \u003cem\u003eJ. Exp. Biol.\u003c/em\u003e, vol. 131, no. 1, pp. 117\u0026ndash;135, Sep. 1987, doi: 10.1242/jeb.131.1.117.\u003c/li\u003e\n\u003cli\u003eS. D. Feldkamp, \u0026ldquo;Foreflipper propulsion in the California sea lion, Zalophus californianus,\u0026rdquo; \u003cem\u003eJ. Zool.\u003c/em\u003e, vol. 212, no. 1, pp. 43\u0026ndash;57, 1987, doi: 10.1111/j.1469-7998.1987.tb05113.x.\u003c/li\u003e\n\u003cli\u003eC. Friedman and M. C. Leftwich, \u0026ldquo;The kinematics of the California sea lion foreflipper during forward swimming,\u0026rdquo; \u003cem\u003eBioinspir. Biomim.\u003c/em\u003e, vol. 9, no. 4, p. 046010, Nov. 2014, doi: 10.1088/1748-3182/9/4/046010.\u003c/li\u003e\n\u003cli\u003eS. D. Feldkamp, R. L. DeLong, and G. A. Antonelis, \u0026ldquo;Diving patterns of California sea lions, Zalophus californianus,\u0026rdquo; \u003cem\u003eCan. J. Zool.\u003c/em\u003e, vol. 67, no. 4, pp. 872\u0026ndash;883, Apr. 1989, doi: 10.1139/z89-129.\u003c/li\u003e\n\u003cli\u003eF. E. Fish, \u0026ldquo;Speed,\u0026rdquo; in \u003cem\u003eEncyclopedia of Marine Mammals\u003c/em\u003e, W.F. Perrin, B. Wursig, and J.G.M. Thewissen., Academic Press, San Diego, 2002, pp. 1161\u0026ndash;1163.\u003c/li\u003e\n\u003cli\u003eD. K. Odell, \u0026ldquo;California sea lion Zalophus californianus,\u0026rdquo; in \u003cem\u003eHandbook of Marine Mammals\u003c/em\u003e, S.H. Ridgway and R. Harrison., vol. 1, Academic Press. London, 1981, pp. 67\u0026ndash;97.\u003c/li\u003e\n\u003cli\u003eS. Godfrey, \u0026ldquo;Additional observations of subaqueous locomotion in the California Sea Lion (Zalophus californianus),\u0026rdquo; \u003cem\u003eAquat. Mamm.\u003c/em\u003e, vol. 11, no. 2, pp. 53\u0026ndash;57, 1985.\u003c/li\u003e\n\u003cli\u003eF. E. Fish, J. Hurley, and D. P. Costa, \u0026ldquo;Maneuverability by the sea lion Zalophus californianus: turning performance of an unstable body design,\u0026rdquo; \u003cem\u003eJ. Exp. Biol.\u003c/em\u003e, vol. 206, no. 4, pp. 667\u0026ndash;674, 2003, doi: 10.1242/jeb.00144.\u003c/li\u003e\n\u003cli\u003eS. Kadapa, A. Drago, N. Marcouiller, J. L. Tangorra, and H. G. Kwatny, \u0026ldquo;Development of numerical model for a bio-inspired sea lion robot,\u0026rdquo; \u003cem\u003eIEEE J. Ocean. Eng.\u003c/em\u003e, 18 2025.\u003c/li\u003e\n\u003cli\u003eN. M. Puzai, A. F. Ayob, and M. R. Arshad, \u0026ldquo;A review on recent advancements in unmanned underwater vehicle design,\u0026rdquo; \u003cem\u003eJ Ocean Mech Aerosp.-Sci Eng\u003c/em\u003e, vol. 31, no. 1, pp. 1\u0026ndash;8, 2016.\u003c/li\u003e\n\u003cli\u003eM.-G. Kim, H. Kang, M.-J. Lee, G. R. Cho, J.-H. Li, and C. Kim, \u0026ldquo;UUV platform optimal design for overcoming strong current,\u0026rdquo; \u003cem\u003eJ. Ocean Eng. Technol.\u003c/em\u003e, vol. 35, no. 6, pp. 434\u0026ndash;445, Dec. 2021, doi: 10.26748/KSOE.2021.069.\u003c/li\u003e\n\u003cli\u003e\u0026ldquo;Underwater Gliders for Ocean Research,\u0026rdquo; ResearchGate. Accessed: Jan. 23, 2025. [Online]. Available: https://www.researchgate.net/publication/233687927_Underwater_Gliders_for_Ocean_Research\u003c/li\u003e\n\u003cli\u003eF. E. Fish, \u0026ldquo;Advantages of aquatic animals as models for bio-inspired drones over present AUV technology,\u0026rdquo; \u003cem\u003eBioinspir. Biomim.\u003c/em\u003e, vol. 15, no. 2, p. 025001, Feb. 2020, doi: 10.1088/1748-3190/ab5a34.\u003c/li\u003e\n\u003cli\u003eA. P. Mignano, S. Kadapa, A. C. Drago, G. V. Lauder, H. G. Kwatny, and J. L. Tangorra, \u0026ldquo;Fish robotics: multi-fin propulsion and the coupling of fin phase, spacing, and compliance,\u0026rdquo; \u003cem\u003eBioinspir. Biomim.\u003c/em\u003e, vol. 19, no. 2, p. 026006, Jan. 2024, doi: 10.1088/1748-3190/ad1dba.\u003c/li\u003e\n\u003cli\u003eA. P. Mignano, S. Kadapa, J. L. Tangorra, and G. V. Lauder, \u0026ldquo;Passing the Wake: Using Multiple Fins to Shape Forces for Swimming,\u0026rdquo; \u003cem\u003eBiomimetics\u003c/em\u003e, vol. 4, no. 1, 2019, doi: 10.3390/biomimetics4010023.\u003c/li\u003e\n\u003cli\u003eC. H. White, G. V. Lauder, and H. Bart-Smith, \u0026ldquo;Tunabot Flex: a tuna-inspired robot with body flexibility improves high-performance swimming,\u0026rdquo; \u003cem\u003eBioinspir. Biomim.\u003c/em\u003e, vol. 16, no. 2, p. 026019, Mar. 2021, doi: 10.1088/1748-3190/abb86d.\u003c/li\u003e\n\u003cli\u003eR. Baines \u003cem\u003eet al.\u003c/em\u003e, \u0026ldquo;Multi-environment robotic transitions through adaptive morphogenesis,\u0026rdquo; \u003cem\u003eNature\u003c/em\u003e, vol. 610, no. 7931, pp. 283\u0026ndash;289, Oct. 2022, doi: 10.1038/s41586-022-05188-w.\u003c/li\u003e\n\u003cli\u003eZ. Chen, T. I. Um, J. Zhu, and H. Bart-Smith, \u0026ldquo;Bio-Inspired Robotic Cownose Ray Propelled by Electroactive Polymer Pectoral Fin,\u0026rdquo; in \u003cem\u003eVolume 2: Biomedical and Biotechnology Engineering; Nanoengineering for Medicine and Biology\u003c/em\u003e, Denver, Colorado, USA: ASMEDC, Jan. 2011, pp. 817\u0026ndash;824. doi: 10.1115/IMECE2011-64174.\u003c/li\u003e\n\u003cli\u003eP. Liljeb\u0026auml;ck and R. Mills, \u0026ldquo;Eelume: A flexible and subsea resident IMR vehicle,\u0026rdquo; in \u003cem\u003eOCEANS 2017 - Aberdeen\u003c/em\u003e, Jun. 2017, pp. 1\u0026ndash;4. doi: 10.1109/OCEANSE.2017.8084826.\u003c/li\u003e\n\u003cli\u003eE. Kelasidi, P. Liljeback, K. Y. Pettersen, and J. T. Gravdahl, \u0026ldquo;Innovation in Underwater Robots: Biologically Inspired Swimming Snake Robots,\u0026rdquo; \u003cem\u003eIEEE Robot. Autom. Mag.\u003c/em\u003e, vol. 23, no. 1, pp. 44\u0026ndash;62, Mar. 2016, doi: 10.1109/MRA.2015.2506121.\u003c/li\u003e\n\u003cli\u003eS. Randeni, E. M. Mellin, M. Sacarny, S. Cheung, M. Benjamin, and M. Triantafyllou, \u0026ldquo;Bioinspired morphing fins to provide optimal maneuverability, stability, and response to turbulence in rigid hull AUVs,\u0026rdquo; \u003cem\u003eBioinspir. Biomim.\u003c/em\u003e, vol. 17, no. 3, p. 036012, Apr. 2022, doi: 10.1088/1748-3190/ac5a3d.\u003c/li\u003e\n\u003cli\u003eY. Wang, Y. Guo, S. Yang, T. Sun, X. Wang, and H. Zhou, \u0026ldquo;Design, Hydrodynamic Analysis, and Testing of a Bio-inspired Movable Bow Mechanism for the Hybrid-driven Underwater Glider,\u0026rdquo; \u003cem\u003eJ. Bionic Eng.\u003c/em\u003e, vol. 20, no. 4, pp. 1493\u0026ndash;1513, Jul. 2023, doi: 10.1007/s42235-023-00361-x.\u003c/li\u003e\n\u003cli\u003eA. M. Leahy \u003cem\u003eet al.\u003c/em\u003e, \u0026ldquo;The role of California sea lion (Zalophus californianus) hindflippers as aquatic control surfaces for maneuverability,\u0026rdquo; \u003cem\u003eJ. Exp. Biol.\u003c/em\u003e, vol. 224, no. 20, p. jeb243020, Oct. 2021, doi: 10.1242/jeb.243020.\u003c/li\u003e\n\u003cli\u003eS. J. Kerr, F. E. Fish, A. J. Nicastro, J. A. Zeligs, S. Skrovan, and M. C. Leftwich, \u0026ldquo;Biomechanical energetics of terrestrial locomotion in California sea lions (Zalophus californianus),\u0026rdquo; \u003cem\u003eJ. Exp. Biol.\u003c/em\u003e, vol. 225, no. 18, p. jeb244163, Sep. 2022, doi: 10.1242/jeb.244163.\u003c/li\u003e\n\u003cli\u003eG. Perrotta, F. E. Fish, D. S. Adams, A. M. Leahy, A. M. Downs, and M. C. Leftwich, \u0026ldquo;Velocity Field Measurements of the California Sea Lion Propulsive Stroke Using Bubble PIV,\u0026rdquo; \u003cem\u003eFluids\u003c/em\u003e, vol. 7, no. 1, Art. no. 1, Jan. 2022, doi: 10.3390/fluids7010003.\u003c/li\u003e\n\u003cli\u003eP. W. Webb, \u0026ldquo;Maneuverability - general issues,\u0026rdquo; \u003cem\u003eIEEE J. Ocean. Eng.\u003c/em\u003e, vol. 29, no. 3, pp. 547\u0026ndash;555, Jul. 2004, doi: 10.1109/JOE.2004.833220.\u003c/li\u003e\n\u003cli\u003eA. Drago, S. Kadapa, N. Marcouiller, H. G. Kwatny, and J. L. Tangorra, \u0026ldquo;Using Reinforcement Learning to Develop a Novel Gait for a Bio-Robotic California Sea Lion,\u0026rdquo; \u003cem\u003eBiomimetics\u003c/em\u003e, vol. 9, no. 9, p. 522, Aug. 2024, doi: 10.3390/biomimetics9090522.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Underwater, Robot, multi-body, Sea lion, Maneuverable, Bio-inspired","lastPublishedDoi":"10.21203/rs.3.rs-7455024/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7455024/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe development of unmanned underwater vehicles (UUVs) capable of operating in complex environments\u0026mdash;such as coastal regions with obstacles and dynamic flows\u0026mdash;requires new and effective maneuvering techniques with high agility to overcome the limitations of current underwater systems. UUVs that can operate in these zones have broad applications, including environmental monitoring, defense, and infrastructure inspection. By studying the swimming and maneuvering strategies of marine organisms, researchers can develop UUVs that integrate biologically inspired characteristics to enhance performance. The California sea lion (\u003cem\u003eZalophus californianus\u003c/em\u003e) was selected as a biological model due to its swimming and maneuvering capabilities in both the open ocean and through the high-energy surf zone. This paper presents the development of a novel, multi-bodied, bio-robotic system with flipper-based propulsion modeled after the California sea lion. An articulatable head and pelvis, flexible fore flippers that generate 3D forces, and adjustable hind flippers were identified as potential contributors to its mobility, as supported by existing research and video analysis. The system serves as a research platform for systematically evaluating how these features influence swimming and maneuvering. Experimental results demonstrate the system's ability to use hydrostatic and hydrodynamic forces to move repeatably in 3D space, providing a foundation for assessing the role of body articulation and flipper movements in underwater locomotion.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e","manuscriptTitle":"Development of a Bio-robotic Swimmer Based on the California Sea Lion","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-11 11:02:33","doi":"10.21203/rs.3.rs-7455024/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d60acfda-3231-4280-b71d-c4de8d78ff9a","owner":[],"postedDate":"September 11th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":54358403,"name":"Biological sciences/Ecology"},{"id":54358404,"name":"Earth and environmental sciences/Ecology"},{"id":54358405,"name":"Physical sciences/Engineering"},{"id":54358406,"name":"Earth and environmental sciences/Ocean sciences"}],"tags":[],"updatedAt":"2026-01-23T06:54:56+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-11 11:02:33","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7455024","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7455024","identity":"rs-7455024","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}