Virtual environmental boundaries reduce cognitive workload for reorientation during turn-by-turn navigation

Virtual environmental boundaries reduce cognitive workload for reorientation during turn-by-turn navigation | 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 Virtual environmental boundaries reduce cognitive workload for reorientation during turn-by-turn navigation Xiaoyu Zhang, Sunao Iwaki This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6491443/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 Current GPS navigation tools primarily help users reach their destinations through turn-by-turn instructions but offer limited support for reorientation, the ability to maintain a sense of direction and self-positioning. Because reorientation during turn-by-turn navigation demands a high cognitive workload, users prioritizing efficiency and safety tend to focus on following instructions rather than encoding their spatial bearings. To address this issue, we proposed visualizing virtual environmental boundaries, such as Augmented Reality (AR) City Walls in the background of the field of view, to serve as a global geometric reference surrounding the navigation area for intuitive direction and distance evaluation during turn-by-turn navigation. Using mobile electroencephalography (EEG), we assessed the cognitive workload of 35 participants as they navigated with virtual reality (VR) headsets. The results indicate that, compared to conventional turn-by-turn navigation using route indications only, displaying environmental boundaries enhances reorientation accuracy while reducing cognitive workload. These findings suggest a potential opportunity for GPS navigation to both the efficiency of reaching a destination and the effectiveness of spatial knowledge acquisition. Biological sciences/Neuroscience/Cognitive neuroscience Biological sciences/Neuroscience/Learning and memory/Spatial memory Biological sciences/Psychology/Human behaviour Physical sciences/Engineering spatial cognition navigation navigation aids cognitive workload virtual reality EEG Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction Current navigation assistants help users reach their destinations efficiently, addressing the challenge of traditional wayfinding, which often involves extensive trial and error. However, after following turn-by-turn instructions (such as turning left at the next intersection), GPS users can reach their destination but still feel disoriented, uncertain of where they are and which direction they are heading. This phenomenon highlights two distinct navigation strategies. The first, route following, requires no encoding of location or direction; instead, users rely on a fixed sequence of responses to specific stimuli, such as turn-by-turn instructions, to reach their destinations { 1} . The second strategy involves spatial knowledge acquisition, referred to as the reorientation strategy, supporting heading recovery and self-location estimation in the allocentric reference frame {2,3,4} . While route following reduces cognitive demands for reaching destinations, it does not support reorientation due to the absence of spatial information encoding, such as directions and locations. In contrast, the reorientation strategy, which involves the spatial knowledge acquisition, contributes to spatial memory and flexible navigation. For example, by encoding the relative location between an initial starting and a goal, the individual can find an alternative route to come back home when a familiar path is blocked {2,3} . Previous research found that users grow increasingly dependent on turn-by-turn navigation, weakening their spatial memory and independent navigation skills {5,6,7} . It raises a question: how do we balance the efficiency of reaching a destination and the effectiveness of spatial knowledge acquisition during GPS navigation? Previous research suggests that current navigation aids struggle to support both strategies simultaneously. Studies on navigation interfaces indicate a trade-off between these two strategies {8,9,10} . A north-up map view, which provides an allocentric reference frame, enables users to quickly judge their location relative to landmarks but requires additional response time to follow turn-by-turn instructions. In contrast, a track-up map or real-street view aligns turn-by-turn route indications with users' egocentric reference frame, facilitating destination-reaching but leading to poor reorientation performance. Although users can mentally switch between egocentric and allocentric reference frames, misalignment between the two spatial reference frames increases response time and error rates of the mental rotation processing {11,12} . Furthermore, navigation aids divert attention from the physical environment, increasing the risk of attention distraction and reducing spatial knowledge acquisition {6,7 ,13 } . These studies suggest that reorientation during turn-by-turn navigation imposes a high cognitive workload. When users prioritize reaching their destination via the shortest and most direct path, they tend to focus on turn-by-turn instructions for convenience, neglecting spatial information acquisition necessary for developing spatial ability. Previous research has attempted to encourage users to adopt spatial-knowledge-related strategies by presenting virtual cues that provide allocentric orientation in real environments. For example, compared to users who followed turn-by-turn instructions, those who navigated using auditory beacons from the direction of the destination demonstrated improved relative location estimation {14} . Similarly, integrating virtual global landmarks that provide stable orientation (such as distant towers and mountains) has been shown to enhance spatial memory compared to navigation guided solely by turn-by-turn instructions {15} . However, in these studies, participants repeatedly encountered the same locations and routes, reinforcing their learning of paths and landmarks through repetition. This reinforcement learning process requires extensive environmental exploration, which contrasts with real-world GPS usage, where users typically minimize exploration, especially when navigating under time constraints, such as during emergency evacuations or urgent business travel in unfamiliar places. It remains unclear whether these methods are effective for reorientation in routes traveled only once, where users cannot rely on reinforcement learning of paths and landmarks. Moreover, previous studies have not demonstrated that virtual cues reduce the cognitive demands of spatial processing during navigation. We suggest that these studies may have underestimated users' preference for minimizing exploration and maximizing navigation efficiency. If virtual orientation cues do not reduce the cognitive demands of spatial knowledge strategies, users may prioritize turn-by-turn instructions. In other words, users are likely to minimize their effort to acquire spatial information if they can reach their destination without it. To address this issue, we argue that navigational aids should both enhance spatial knowledge acquisition and reduce the associated cognitive workload. In our previous research, we proposed visualizing virtual environmental boundaries as a global geometric reference that encloses the navigation area, providing intuitive support for direction and distance evaluation during turn-by-turn navigation {16} . Specifically, we introduced Augmented Reality (AR) City Walls, a design prototype that overlays encircling walls in the background of the user's field of view, positioned along the perimeter of a square navigation area defined by feasible routes between the starting point and the destination (see Fig. 1). The design was inspired by geometric module theory, which suggests that animals possess an innate cognitive mechanism to process spatial information based on environmental geometry, such as walls of rooms and shapes of navigation area {17,18,19} . For example, rodents use specific sides or corners of a chamber's geometric boundary to locate the hidden food or judge their position relative to the boundary {20,21,22} . Human studies similarly support the role of boundaries in orientation {20,23,24} and relative location evaluation {25,26} . Although our previous study had found that incorporating turn-by-turn indication with AR City Walls can improve reorientation performance without extensive exploration of environment {16} , it remains unclear whether environmental boundaries can reduce the cognitive workload associated with reorientation during turn-by-turn navigation. Since high-visibility geometric cues are rare in large-scale outdoor environments, no prior research has objectively measured the cognitive workload involved in using such cues within turn-by-turn navigation. In this study, we examine the effects of environmental boundaries (i.e., AR City Walls) on reorientation performance at the arrival of novel locations and the cognitive workload for using them during turn-by-turn navigation. We hypothesize that the virtual environmental boundaries can improve reorientation performance and reduce the cognitive workload when users follow a predefined route in unfamiliar environments, thereby facilitating efficient spatial knowledge acquisition while preserving the benefits of turn-by-turn navigation. We emphasize that both improved reorientation performance and reduced cognitive load are necessary conditions. If either condition is not met, environmental boundaries cannot efficiently improve spatial cognition during turn-by-turn navigation. To test this hypothesis, we used mobile electroencephalography (EEG) to record brain activity while participants navigated simulated large-scale outdoor environments using virtual reality (VR) headsets. The VR headsets provided a realistic and expansive pedestrian perspective, allowing participants to control their movement direction through head rotations {27,28} . We recruited 42 participants, each experiencing two conditions (i.e., Non-boundary and Boundary). After excluding individuals affected by VR sickness and EEG signal artifacts, 35 participants remained for the final analysis. To minimize the influence of extensive exploration and ensure sufficient data sampling, each participant navigated both conditions twice in different environments. Figure 2 shows one of the environments used in the experiment. The environment was visually occluded by dense forest, preventing visibility of other locations or distinctive landmarks. Each environment contained 25 specific location areas, with one centrally located 'Home' serving as both the initial starting point and the destination of the journey. A circular, winding main road connected these areas, ensuring that each location could be visited only once, except for Home. No identifying signs or landmarks were displayed at any location area, making each area visually indistinct from the surrounding road (pink and green areas surrounded by yellow circles in Fig. 2 for illustration purposes only and were not visible in the VR environment). Participants were informed that 'Home' was their initial standing point, which was randomized for each environment to encourage self-guided spatial encoding. To induce disorientation, the roads between location areas included one or two U-turns. In both conditions, participants were instructed to face Home upon entering a new location area. The angular difference between their response and the actual direction of Home was recorded as a measure of reorientation performance. After responding, an arrow appeared to indicate the next travel direction, simulating turn-by-turn navigation. The arrow disappeared once participants exited the location area. In the Boundary condition, AR City Walls were displayed at two key points: before the start of travel, to aid initial orientation, and after the exit of location areas, to support reorientation. The walls disappeared after the left of the initial start point or the entry of the location areas. These designs limited continuous exposure to AR City Walls and required participants to rely on spatial knowledge acquired during navigation to recall Home's direction. We used the task-irrelevant auditory probe technique to assess objective cognitive workload for several reasons. This classical EEG paradigm estimates the attentional resources allocated to a task, particularly for complex and sustained activities such as driving {29,30} and video watching {33} . Compared to subjective self-reports {29,33} , such as the NASA-TLX questionnaire {32} , it provides a more accurate and objective assessment of the cognitive demands required for task completion without disrupting execution, as the task relies on non-auditory sensory modalities. Based on the principle that the available mental resources are limited at a given time {33} , participants passively listened to a series of auditory probes while performing target tasks. Event-related potentials (ERPs) such as N100 and P300 components were triggered by those task-irrelevant auditory probes, the amplitudes of which decrease as the cognitive workload of the target task increases {29,33,34} . In our experiment, auditory probes were presented continuously as background sounds. Only EEG epochs between the participant's departure from the previous location area and arrival at the next location area were selected for analysis, as this period corresponded to the display of the AR City Walls, the sole difference between the settings of Non-boundary and Boundary conditions. The mean N100 amplitude triggered by these probes at the FCz channel was calculated, as it is the most sensible component to assess objective cognitive workload according to previous research {29,33} . Additionally, participants completed the NASA-TLX questionnaire after navigation to provide subjective workload measures. Together, N100 mean amplitudes and NASA-TLX scores provided a comprehensive and comparable evaluation of cognitive workload. In summary, we hypothesized that the presence of virtual environmental boundaries would enhance reorientation performance and reduce cognitive workload during turn-by-turn navigation. 2. Results Given our within-subject design (Non-boundary vs. Boundary), the two-tailed paired-sample t-test with a 95% confidence interval was used for statistical analysis. 2.1. Environmental boundary improves reorientation performance Reorientation performance was assessed by calculating the average angular difference between the actual direction of Home and the direction participants faced when instructed to orient toward it. Results showed that participants were significantly more accurate in estimating their relative position to Home in the Boundary condition (M = 19.98, SD = 10.68) compared to the Non-boundary condition (M = 57.96, SD = 16.37), t(34) = -15.00, p < 0.001, CI[-43.13, -32.84] (see Fig. 3). These findings suggest that the presence of environmental boundaries significantly enhances reorientation performance during turn-by-turn navigation without requiring extensive environmental exploration. Notably, since the AR City Walls were not displayed while participants navigated the location areas or made directional responses, the observed improvement in reorientation accuracy can be attributed to intermittent rather than continuous exposure to the boundaries. 2.2. Environmental boundary reduce the cognitive workload for reorientation 2.2.1. Objective cognitive workload measured by N100 mean amplitude Figure 4 presents the grand average event-related potentials (ERPs) at electrode FCz for both conditions across all participants. The peak of the N100 component is visibly higher in the Non-boundary condition compared to the Boundary condition. Mean N100 amplitude was calculated within the 78–114 ms time window. Results showed that the Boundary condition (M = -2.20, SD = 1.27) had a significantly lower mean amplitude than the Non-boundary condition (M = -1.86$, SD = 1.15), t(34) = -3.79, p = 0.001, CI[-0.53, -0.16] (see Fig. 5). Since lower (i.e., more negative) N100 amplitudes are associated with reduced cognitive workload, this result indicates that the Boundary condition imposed less cognitive demand than the Non-boundary condition. 2.2.2. Subjective workload measured by NASA-TLX scores Subscale weightings were applied in NASA-TLX to compute overall workload scores. Since the only difference between the Boundary and Non-boundary conditions was the perception of AR City Walls, we interpret the overall NASA-TLX score as an indicator of cognitive demand. Higher scores reflect greater cognitive workload during the GPS navigation task. Results showed that the Boundary condition (M = 50.68, SD = 14.79) was associated with significantly lower cognitive workload than the Non-boundary condition (M = 74.15, SD = 8.21), t(34) = -8.91, p < 0.001, CI[-28.83, -18.12] (see Fig. 6). This subjective workload measure aligns with the objective cognitive workload indexed by the N100 mean amplitude. Together, these findings suggest that displaying environmental boundaries reduces the cognitive demands associated with reorientation during turn-by-turn navigation. 3. Discussion This study aimed to investigate the effect of virtual environmental boundaries on the efficiency of spatial knowledge acquisition. We measured reorientation performance and assessed cognitive workload for spatial knowledge acquisition using EEG data during navigation, and a self-report questionnaire was administered afterward. In a within-subjects design, 42 participants navigated unfamiliar environments under both non-boundary and boundary conditions. To prevent extensive exploration of environments, each participant followed a predetermined route, ensuring that they were not exposed to the same locations or paths multiple times. We hypothesized that virtual environmental boundaries would enhance spatial information acquisition during turn-by-turn navigation if both criteria were met: (1) improved reorientation accuracy and (2) reduced cognitive workload during navigation. Our results indicate that displaying virtual environmental boundaries enhances reorientation performance and reduces cognitive workload during turn-by-turn navigation. Participants demonstrated greater accuracy in estimating the relative location of Home after traveling with environmental boundaries for part of the route. This improvement in reorientation performance was accompanied by a decrease in cognitive workload, as reflected in the N100 mean amplitude and NASA-TLX scores. These findings suggest that virtual environmental boundaries can support efficient spatial information acquisition while preserving the benefits of turn-by-turn navigation in reaching a destination. Our experimental setting was intentionally designed to avoid repeated exposure to the same locations or paths. As a result, the observed improvement in spatial knowledge acquisition can be primarily attributed to the presence of virtual environmental boundaries rather than interactions with other environmental cues. This implies that the environmental boundary can play a prominent role in spatial knowledge acquisition during large-scale outdoor navigation. The geometric module theory provides a plausible explanation for the role of virtual environmental boundaries in spatial processing: the global geometric layout of environmental boundaries offers intuitive direction and distance evaluation within a stable allocentric reference frame {17,18,19} . Unlike navigation aids based on map views {8,9,11,12} , which require substantial effort to align egocentric and allocentric reference frames, environmental boundaries may facilitate an automatic transformation between these spatial reference frames, thereby reducing the cognitive workload associated with reorientation. Unlike previous research that assessed spatial knowledge after reinforcement learning of paths and landmarks {14,15} , our study focused on spatial knowledge acquisition during one-time navigation. This approach provides a more accurate evaluation of the efficiency of navigation aids in spatial information acquisition. This study highlights that users tend to maximize benefits while minimizing costs. When the primary goal of navigation is to reach a destination, if the cognitive workload required for spatial information acquisition is too high and spatial knowledge is not necessary for reaching the goal, users may choose to disregard spatial-knowledge-related navigation strategy. Based on this premise, this study leveraged EEG, specifically the N100 component, to provide an objective neural measure of cognitive workload during navigation. Compared to traditional self-report methods, EEG offers high temporal resolution and can capture subtle, real-time fluctuations in cognitive processing that may not be consciously perceived by participants. By integrating both subjective (NASA-TLX) and objective (N100 amplitude) data, our approach enhances the validity of the findings and demonstrates the value of neurophysiological methods in evaluating spatial cognition. This methodology offers a powerful tool for future research aiming to understand and improve the cognitive efficiency of navigation systems. In brief, we suggest that enhanced reorientation performance increases the benefits of navigation, while a reduced cognitive workload lowers the cost, allowing users to naturally utilize virtual environmental boundaries to improve their spatial knowledge while retaining the advantage of turn-by-turn instructions for route following. However, like many previous studies, our research explicitly instructed participants to perform spatial cognition tasks (e.g., pointing tasks, relative location estimation), which may have encouraged active acquisition and processing of spatial information. Future studies should examine spatial processing efficiency in real-world field settings, where users engage in typical navigation behaviors using GPS systems without explicit instructions to acquire spatial knowledge. Additionally, we did not assess long-term spatial memory or cognitive map representation after complete learning. Future research should explore both the efficiency of spatial encoding and its long-term impact on cognitive map development. Furthermore, as our experiment simulated pedestrian navigation in a forested environment with sparse environmental cues and virtual reality constraints on participant movement, the impact of environmental boundaries should be evaluated in a broader range of real-world scenarios, such as driving and urban navigation, for practical applications. 4. Materials and Methods 4.1. Participants Forty-two individuals with normal or corrected-to-normal vision were recruited via a local online bulletin board. The sample included 20 women (mean age = 25.7 years, SD = 2.6, range = 21–32). All participants were right-handed and provided written informed consent before the study. Ethical approval was obtained from the local ethics committee (Internal Review Board, National Institute of Advanced Industrial Science and Technology, Japan; Protocol No. 2019-481). All experiments were conducted in accordance with relevant institutional guidelines and regulations. Due to simulator sickness and EEG data artifacts, seven participants were excluded, resulting in a final sample of 35. All participants followed the same experimental procedure. 4.2. Equipment 4.2.1. Virtual r eality HTC Vive Pro Eye head-mounted display (HTC Corporation, Taoyuan, Taiwan) with a 90 Hz refresh rate (Dual OLED 3.5" diagonal screen, 1440 × 1600 pixels per eye, 120 Hz gaze data output frequency, 615 ppi, and 110° nominal field of view) was used to present the virtual environment. The VR headset was equipped with a mobile EEG device. The EEG amplifier and data transmission unit were carried in a backpack (see Fig 2a). For more details about EEG equipment, see the section about EEG recordings and analysis. The VR setup was placed in a 3.5 m × 3.5 m room with a ceiling height of 2.5 m, while the virtual environment spanned an area of 500 m × 500 m. To prevent movement obstruction, the VR headset cable was suspended from the center of the ceiling. A VR motion tracker monitored participants' head position and orientation. The movement of participants was a combination of the head rotation recorded by the VR movement tracker and trackpad on the left-hand controller. The positive and negative y-axes of the trackpad shifted forward and backward, respectively. The movement speed (1.2 ± 1.2 m/s) increased from the middle to both ends of the axis. The direction of movement was aligned with the heading in a 2-dimension plane parallel to the transverse section of the individual head. The back-and-forth pan, controlled by the left trackpad and multiplied by the head rotation, resulted in smooth movement. This movement prevents the occurrence of simulation sickness induced by bouncing up and down during actual walking. The trigger button on the right-hand controller was used to interact with the visual instruction window, a 2-dimensional panel right in front of the view of participants. 4.2.2. Map layouts Four maps for the navigation experiment were built using the Unity3D game engine (Unity Technologies, San Francisco, California, USA): one for practice and the other three for the formal experiment (for an example, see Fig. 2). The navigation area of each map was 500 m × 500 m. A winding, closed-loop road passing through the forest formed the main road of each map. The main road in the practice map passed through 19 location areas. In the formal test, each of the three main roads traversed 25 location areas, including a central 'Home' location that served as both the starting point and the final destination. Thirteen of these location areas were configured as nested hexagons on each map (rendered in green in Fig. 2b). These two hexagons had the same center, which was the home. The side length (240 m) of the first hexagon was twice that of the second. The sequence of roads connecting these 13 fixed locations varied across maps. The remaining location areas (6 in the practice map and 12 in each formal test map, rendered in pink in Fig. 2b) were positioned after connecting the 13 fixed points with a circular road. These additional location areas were placed based on the condition that one or two turns must occur between any two consecutively visited areas on the main route. The main roads lacked clear geometric information, such as right angles and straight lines. The roads were intentionally designed without clear geometric features such as right angles or straight paths, to enhance disorientation between successive locations. Participants' views were obstructed by dense forest, preventing them from seeing other location areas. Apart from trees and vegetation, no significant landmarks were present. Moreover, no identifying signs or markers were displayed at any of the location areas, making each one visually indistinguishable from other parts of the road. While each location area was defined by a 12-meter radius circle, this circle was not visible in the VR environment. 4.2.3. Navigation aids Participants navigated the virtual environment using different navigation aids depending on the condition: Boundary or Non-boundary. Turn-by-turn indication (used in both conditions) In both conditions, participants received turn-by-turn navigation assistance. A virtual arrow appeared at each location area, indicating the next forward direction. Participants were instructed to follow the arrow to reach the next location. Once they passed over the location area, the arrow disappeared. The arrow was not displayed during travel along the road between two location areas. AR City Walls (used only in the boundary condition) The AR City Walls served as a virtual environmental boundary and acted as a high-visibility geometric cue providing an allocentric reference frame for orientation and distance estimation. This boundary consisted of four virtual city walls placed in the cardinal directions (i.e., North, South, East, and West) enclosing the square-shaped navigation environment. The distinct visual features of each wall enabled participants to identify the cardinal directions. The north wall featured a blue base with a central spire-topped building, while the south wall had a similar structure but was gray. The east and west walls were distinguished by white bases and central towers, red on the east and blue on the west (see Fig. 2b). When participants stood at the center of the environment facing north, the northern city wall with the blue spire was directly in front of them. As they moved northward, the northern wall appeared progressively larger and more prominent in their field of view (see Fig. 1c), reinforcing their perception of direction and distance within the environment. 4.2.4. Task-irrelevant auditory probe Task-irrelevant auditory probes were displayed as the background sounds in each run and programmed using MATLAB (MathWorks, Inc.). Similar to previous studies {29,30,33} , the probe sequence consisted of 12 pure tones ranging from 500 to 1600 Hz at intervals of 100 Hz. The duration of each probe was 50 ms, including rise and fall times of 10 ms. All tones were presented randomly with equal probability under the precondition that probes with the same frequency did not appear successively. The stimulus interval was randomized in the range of 400–800 ms (mean: 600 ms). The probe sequence was provided at approximately 75 dB/SPL using earphones on the VR headset. Participants were asked to neglect the background sounds, and none reported discomfort due to the probe presentation. 4.3. Experimental procedure 4.3.1. Navigation During the experiment, participants were never exposed to a global map or an aerial view of the navigation area. A run was defined as a full traversal of the main road on a given map, returning to the starting point. Participants completed two runs during the practice session using the same map, and four runs during the formal test using three different maps. Each run followed an identical procedure; however, the practice session was designed to be less challenging than the formal test. Therefore, we describe the formal test procedure first. In the formal test, each condition (Boundary and Non-boundary) consisted of two runs, resulting in a total of four runs per participant. A fully randomized design ensured that no condition was paired with the same map layout more than once, and that the same condition and layout were not repeated consecutively. The condition–layout combinations and run order were counterbalanced across participants to control for order effects. Each run lasted approximately 20 minutes. After completing each navigation run, participants removed the VR headset, filled out the NASA-TLX questionnaire on a laptop using mouse input, and then took a short break. At the beginning of each run, participants started at Home, the central point of the navigation area. While at Home, participants were stationary but could freely rotate their heads to encode their initial position. In both conditions, the first directional arrow indicating the starting direction was visible. In the Boundary condition, participants could also perceive the AR City Walls at this stage. Participants were informed that the Home location was randomized for each run, and none of them knew its position in advance. Once participants felt confident, they had encoded the Home location, they pressed the right-hand controller trigger to begin the navigation. Upon confirmation, the white directional arrow turned blue in both conditions, and the AR City Walls disappeared in the Boundary condition. Each time participants navigated from one location area to the next, a trial was completed. Accordingly, each formal test run included 24 trials. Each trial followed the same procedure. After leaving the current location area, the directional arrow at that location disappeared in both conditions. In the Boundary condition, the AR City Walls were re-displayed only after the participant exited the current location area, while no additional navigation aids were provided in the Non-boundary condition. Upon entering the next location area, participants were again stationary but allowed to rotate their heads to observe the surroundings. At this point, the AR City Walls (if visible) automatically disappeared. Then, a visual instruction window appeared, prompting participants to face toward the Home location, which was not visible. Participants indicated their estimated direction to Home by physically turning to face it and pressing the right trigger to confirm their response. No feedback was given about the accuracy of their estimation. Reorientation performance was assessed by calculating the angular difference between the participant's facing direction and the actual direction to Home. Next, a visual instruction reminded participants to observe the directional arrow, which appeared after they confirmed their Home direction. After pressing the right-hand trigger again, the instruction window disappeared, and the arrow turned blue, signaling that they could proceed to the next location. If participants moved in the wrong direction, a warning window appeared in front of them, guiding them back to the previous location area. This process was repeated until they completed the full loop and returned to Home. The road length in the practice sessions was shorter than in the formal experiment, with 19 location areas in the practice map compared to 25 in each of the formal test layouts. During the practice runs, navigation aids, including the directional arrows and AR City Walls, were continuously displayed from start to finish in both runs. This approach was intended to help participants become familiar with the navigation aids and task procedures. Participants were allowed to perceive the AR City Walls when estimating the Home direction during practice. They could also stop the practice session at any time, and completion of the NASA-TLX questionnaire was not required. Data collected during the practice session were excluded from the final analysis. 4.3.2. EEG recordings and analysis EEG activity was recorded using the BrainAmp Standard amplifier, the BrainVision MOVE 32-channel wireless transmission system, and the actiCAP slim active EEG electrode system (Brain Products GmbH, Germany). Data analysis was performed using the EEGLAB toolbox {35} in MATLAB. During navigation, the transmission system was secured in a backpack worn by participants. The EEG system used 28 electrode sites (Fp1, Fp2, F7, F3, Fz, F4, F8, FCz, T7, C3, Cz, C4, T8, TP7, CP3, CPz, CP4, TP8, P7, P3, Pz, P4, P8, POz, O1, Oz, O2, and M2 in the extended 10–20 System), with M1 and AFz as the reference and ground electrodes, respectively. Horizontal and vertical electrooculograms (EOG) were recorded using electrodes placed on the outer left and right canthi and above and below the right eye. The impedance of all electrodes was maintained below 10 kΩ. EEG signals were digitized at a sampling rate of 1000 Hz, and the time constant was set to 10 s. Preprocessing was conducted to clean the raw EEG data and increase the signal-to-noise ratio of the movable EEG datasets. The EEG signals were re-referenced to the averaged EEG signals from the linked mastoids (M1 and M2) and band-pass filtered at 0.1–30 Hz. Artifacts derived from the EOG and body movements were removed from the data using independent component analysis {35} . To compute ERPs elicited by probe stimuli during the reorientation stage, averaged epochs were extracted from the EEG segmentation of the specific stage with a data window of 500 ms, including a 100 ms pre-stimulus baseline period. Prior to ERP averaging, epochs in which the EEG signal variation exceeded ± 80 μV were removed. In each condition, epochs from the two runs were combined to ensure adequate data sampling after artifact rejection (the mean remaining epoch numbers for the Boundary and Non-boundary conditions were 630 and 650, respectively). As reported in previous research {29,33} , the mean N100 amplitude at FCz is the most sensitive indicator of cognitive workload compared to that at the other electrode sites. Therefore, the average ERPs at the FCz were calculated. The grand average ERP at FCz across 35 participants showed two negative peaks around 100 ms, likely due to variability in individual ERP latencies (see Fig. 4). However, an examination of each participant's ERP in each condition revealed only one maximum negative peak. To determine the appropriate time window for calculating the mean amplitude of N100, we first computed the mean (96 ms) and standard deviation (18 ms) of the N100 peak latency. Based on this, the mean amplitude of N100 was calculated as the mean amplitude in the time window of 78–114 ms. The N100 mean amplitude assessed the objective cognitive workload during navigation. 4.4. Statistical analysis Statistical analysis was conducted using SPSS 23 (IBM, Armonk, NY, USA). A two-tailed paired-sample t-test with a 95% confidence interval was used to compare conditions. Three key measures were analyzed: Reorientation performance , assessed by the angular difference between the actual direction of Home and the direction participants faced when instructed to orient toward Home. The average reorientation performance across 48 trials (combining the first and second periods) was calculated. Objective cognitive workload , measured by the mean amplitude of the N100 component. For each participant in each condition, the N100 mean amplitude was calculated within a time window of 78–114 ms. Subjective cognitive workload , assessed using NASA-TLX scores. In each run, NASA-TLX scores were computed as a weighted average of six subscales: mental demand, physical demand, temporal demand, performance, effort, and frustration. The average overall NASA-TLX score across the first and second periods was calculated for each condition. Since the perception of AR City Walls was the only difference between the Boundary and Non-boundary conditions, we suggest that the overall NASA-TLX score serves as a measure of cognitive demand. The Shapiro-Wilk test was used to assess the normality of all data samples. The results indicated that subjective and objective workload data were normally distributed. Although reorientation performance data did not follow a normal distribution, outliers were not removed to avoid artificially inflating the significance of the results (see Fig. 3). Declarations Acknowledgements We thank Ms. Shinobu Yasumuro at the National Institute of Advanced Industrial Science and Technology and Ms. Weiran Cui at the University of Tsukuba for their support in coordinating the experiment. This study was supported in part by the Japan Society for the Promotion of Science (JSPS) KAKENHI [grant numbers 21H03787 and 23K21851] and JST SPRING [grant number JPMJSP2124]. Author contributions statement XY.Z and S.I conceptualized the study and conducted data acquisition. XY.Z developed the methodology, performed data analysis, wrote the main manuscript, and prepared the figures. Both XY.Z and S.I reviewed and edited the manuscript. Data availability statement The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request. Additional information Competing interests ： The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. References Cook, D. & Kesner, R. P. Caudate nucleus and memory for egocentric localization. Behav. Neural Biol . 49 , 332-343, DOI: 10.1016/S0163-1047(88)90338- X (1988). Tolman, E. C. Cognitive maps in rats and men. Psychol. Rev . 55 , 189Œ208, DOI:10.1037/h0061626 (1948). Epstein, R. A., Patai, E. Z., Julian, J. B. & Spiers, H. J. The cognitive map in humans: spatial navigation and beyond. Nat. Neurosci . 20 , 1504-1513, DOI:10.1038/nn.4656 (2017). Julian, J. B., Keinath, A. T., Marchette, S. A. & Epstein, R. A. The neurocognitive basis of spatial reorientation. Curr.Biol . 28 , R1059-1073, DOI:10.1016/j.cub.2018.04.057 (2018). Dahmani, L. & Bohbot, V. D. Habitual use of gps negatively impacts spatial memory during self-guided navigation. Sci.Reports 10 , 6310, DOI:10.1038/s41598- 020- 62877- 0 (2020). Gardony, A. L., Brunyé, T. T., Mahoney, C. R. & Taylor, H. A. How navigational aids impair spatial memory: Evidence for divided attention. Spatial Cognition. & Comput . 13 , 319-350, DOI:10.1080/13875868.2013.792821 (2013). Ying, Q., Dong, W. & Fabrikant, S. I. How do in-car navigation aids impair expert navigators spatial learning ability? Annals Am. Assoc. Geogr . 114 , 1483-1504, DOI:10.1080/24694452.2024.2356858 (2024). Münzer, S., Zimmer, H. D. & Baus, J. Navigation assistance: a trade-off between wayfinding support and configural learning support. J. experimental psychology: applied 18 , 18-37, DOI:10.1037/a0026553 (2012). Münzer, S., Lörch, L. & Frankenstein, J. Wayfinding and acquisition of spatial knowledge with navigation assistance. J. Exp. Psychol. Appl . 26 , 73-88, DOI:10.1037/xap0000237 (2020). Qiu, X., Yang, Z., Yang, J., Wang, Q. & Wang, D. Impact of ar navigation display methods on wayfinding performance and spatial knowledge acquisition. Int. J. Human Computer Interact . 40 , 2676-2696, DOI:10.1080/10447318.2023.2169524 (2024). Aretz, A. J. & and, C. D. W. The mental rotation of map displays. Hum. Perform . 5 , 303-328, DOI:10.1207/s15327043hup0504_3 (1992). Shepard, R. N. & Hurwitz, S. Upward direction, mental rotation, and discrimination of left and right turns in maps. Cognition 18 , 161-193, DOI:10.1016/0010- 0277(84)90024- 6 (1984). Ishikawa, T., Fujiwara, H., Imai, O. & Okabe, A. Wayfinding with a gaps-based mobile navigation system: A comparison with maps and direct experience. J. Environ. Psychol . 28 , 74-82, DOI:10.1016/j.jenvp.2007.09.002 (2008). Clemenson, G. D., Maselli, A., Fiannaca, A. J., Miller, A. & Gonzalez-Franco, M. Rethinking gps navigation: creating cognitive maps through auditory clues. Sci. Reports 11 , 7764, DOI:10.1038/s41598- 021- 87148- 4 (2021). Liu, J., Singh, A. K., Wunderlich, A., Gramann, K. & Lin, C.-T. Redesigning navigational aids using virtual global landmarks to improve spatial knowledge retrieval. npj Sci. Learn . 7 , 17, DOI:10.1038/s41539- 022- 00132- z (2022). Zhang, X. & Iwaki, S. Self-location and reorientation of individuals without reading maps: Increased spatial memory during gps navigation using ar city walls. In Wei, J. & Margetis, G. (eds.) Human-Centered Design, Operation and Evaluation of Mobile Communications , 254-267, DOI:10.1007/978- 3- 031- 60458- 4_17 (Springer Nature Switzerland, Cham, 2024). Cheng, K. A purely geometric module in the rat's spatial representation. Cognition 23 , 149-178, DOI:10.1016/0010- 0277(86)90041- 7 (1986). Tommasi, L., Chiandetti, C., Pecchia, T., Sovrano, V. A. & Vallortigara, G. From natural geometry to spatial cognition. Neurosci. & Biobehav. Rev. 36 , 799-824, DOI:10.1016/j.neubiorev.2011.12.007 (2012). Cheng, K. & Newcombe, N. S. Is there a geometric module for spatial orientation? squaring theory and evidence. Psychon. Bull. & Rev . 12 , 1-23, DOI:10.3758/BF03196346 (2005). Keinath, A. T., Julian, J. B., Epstein, R. A. & Muzzio, I. A. Environmental geometry aligns the hippocampal map during spatial reorientation. Curr. Biol. 27 , 309-317, DOI:10.1016/j.cub.2016.11.046 (2017). Weiss, S.et al. Consistency of spatial representations in rat entorhinal cortex predicts performance in a reorientation task. Curr. Biol . 27 , 3658-3665, DOI:10.1016/j.cub.2017.10.015 (2017). Lever, C., Burton, S., Jeewajee, A., O'Keefe, J. & Burgess, N. Boundary vector cells in the subiculum of the hippocampal formation. J. Neurosci. 29 , 9771-9777, DOI:10.1523/JNEUROSCI.1319- 09.2009 (2009). Shine, J. P., Valdés-Herrera, J. P., Hegarty, M. & Wolbers, T. The human retrosplenial cortex and thalamus code head direction in a global reference frame. J. Neurosci. 36 , 6371-6381, DOI:10.1523/JNEUROSCI.1268- 15.2016 (2016). Marchette, S. A., Vass, L. K., Ryan, J. & Epstein, R. A. Anchoring the neural compass: coding of local spatial reference frames in human medial parietal lobe. Nat. Neurosci. 17 , 1598-1606, DOI:10.1038/nn.3834 (2014). Bellmund, J. L. S.et al. Deforming the metric of cognitive maps distorts memory. Nat. Hum. Behav . 4 , 177-188, DOI:10.1038/s41562- 019- 0767- 3 (2020). Keinath, A. T., Rechnitz, O., Balasubramanian, V. & Epstein, R. A. Environmental deformations dynamically shift human spatial memory. Hippocampus 31 , 89-101, DOI:10.1002/hipo.23265 (2021). Taube, J. S., Valerio, S. & Yoder, R. M. Is navigation in virtual reality with fmri really navigation? J. Cogn. Neurosci . 25 ,1008-1019, DOI:10.1162/jocn_a_00386 (2013). Makeig, S., Gramann, K., Jung, T.-P., Sejnowski, T. J. & Poizner, H. Linking brain, mind and behavior. Int. J . Psy-chophysiol . 73 , 95-100, DOI:10.1016/j.ijpsycho.2008.11.008 (2009). Neural Processes in Clinical Psychophysiology. Sugimoto, F.et al. Effects of one-pedal automobile operation on the driver's emotional state and cognitive workload. Appl. Ergonomics 88 , 103179, DOI:10.1016/j.apergo.2020.103179 (2020). Takeda, Y., Inoue, K., Kimura, M., Sato, T. & Nagai, C. Electrophysiological assessment of driving pleasure and difficulty using a task-irrelevant probe technique. Biol. Psychol . 120 , 137-141, DOI:10.1016/j.biopsycho.2016.09.009 (2016). Takeda, Y. & Kimura, M. The auditory n1 amplitude for task-irrelevant probes reflects visual interest. Int. J. Psychophysiol. 94 , 35-41, DOI:10.1016/j.ijpsycho.2014.07.007 (2014). Hart, S. G. & Staveland, L. E. Development of nasa-tlx (task load index): Results of empirical and theoretical research.In Hancock, P. A. & Meshkati, N. (eds.) Human Mental Workload , vol. 52 of Advances in Psychology , 139-183, DOI:10.1016/S0166- 4115(08)62386- 9 (North-Holland, 1988). Kahneman, D. Attention and effort, vol. 1063 (Prentice-Hall, 1973). Kramer, A. F., Trejo, L. J. & Humphrey, D. Assessment of mental workload with task-irrelevant auditory probes. Biol.Psychol. 40 , 83-100, DOI:10.1016/0301- 0511(95)05108- 2 (1995). EEG in Basic and Applied Settings. Delorme, A. & Makeig, S. Eeglab: an open source toolbox for analysis of single-trial eeg dynamics including independent component analysis. J. Neurosci. Methods 134 , 9-21, DOI:10.1016/j.jneumeth.2003.10.009 (2004). Additional Declarations No competing interests reported. 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-6491443","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":466217718,"identity":"7ede2787-767c-42cd-9444-d6217ef1e068","order_by":0,"name":"Xiaoyu Zhang","email":"","orcid":"","institution":"University of Tsukuba","correspondingAuthor":false,"prefix":"","firstName":"Xiaoyu","middleName":"","lastName":"Zhang","suffix":""},{"id":466217719,"identity":"9cc53019-4818-4445-a3a9-0497ea9bdbc9","order_by":1,"name":"Sunao Iwaki","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2UlEQVRIie2RPwrCMBSH36NDlxwggrRXaHBx8DDt0k69gmRKl6JrQS/hDQoBXQquFZeKg4tDx6IdjAoiLombSL7twfv4vT8AFstPghygpB44b7WRMlLKs9lAuVNCxI2bfe4I2lXjZJ65pwYuE3AWGjMoUQzymqaFJIzjLAZclhoFUNSkpSmXBDnmErAIdYOh2PUtTXzpHswUUIPtSU3DQALj0BkogcTsOqwoW6ldiojHRLuLn2VHdl5PfW+7adq2n3hMd7HX0x9EAlSWxvigV7n0O8VisVj+nxvli0ahcBglRQAAAABJRU5ErkJggg==","orcid":"","institution":"National Institute of Advanced Industrial Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Sunao","middleName":"","lastName":"Iwaki","suffix":""}],"badges":[],"createdAt":"2025-04-21 01:08:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6491443/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6491443/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":84201988,"identity":"f9da7f8e-5dee-4dc9-a033-4267469b694a","added_by":"auto","created_at":"2025-06-09 08:37:24","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":139607,"visible":true,"origin":"","legend":"\u003cp\u003eWorking principle of AR City Walls. (a) Steps for generating a virtual environmental boundary: First, the navigation system plans the path to a series of locations. Second, a square navigation area is defined by feasible routes. Third, the AR City Walls are positioned along the perimeter. (b) In a real-world setting, the AR City Walls appear in the user's field of view without obstructing the surrounding street view. A tower with a blue triangular roof is positioned at the center of the northern city wall. (c) In our virtual reality experiment simulating a forested environment, when users reached the locations marked in panel (a) and faced north, they estimated their relative location and orientation based on their view of the AR City Walls.\u003c/p\u003e","description":"","filename":"Figure1ARCityWall.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6491443/v1/7dd31779abb012ad4253e529.jpg"},{"id":84201987,"identity":"f43907f5-e4ef-44fa-aeaa-77b429bec50a","added_by":"auto","created_at":"2025-06-09 08:37:24","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":118743,"visible":true,"origin":"","legend":"\u003cp\u003eExperimental environment. (a) Participants in our experiment were equipped with virtual reality headsets and mobile EEG devices, with the EEG amplifier and data transmission unit carried in a backpack. The virtual environment simulated forest roads, with low-polygon plants instead of realistic vegetation, to minimize the risk of virtual reality sickness. In the Non-boundary condition, navigation aids consisted of virtual arrows providing directional guidance for route-following. In the Boundary condition, navigation aids included directional arrows and the AR City Walls. (b) An aerial view of one of the three experimental layouts, surrounded by illustrated front elevations of the AR City Walls. A distinguishing building is located at the center of each side of the city walls. In each layout, participants navigated through a forest containing 25 locations connected by a circular main road. Thirteen locations (green) remain fixed across all layouts, while the remaining 12 locations (pink) are randomized in each of the three layouts. The center of each map represents Home, where participants both began and ended their navigation. The blue arrow indicates the initial direction of travel. Trees and other vegetation within and around the environment are removed in the aerial view to provide an unobstructed visualization of the main road.\u003c/p\u003e","description":"","filename":"Figure2Environment.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6491443/v1/64486ffad03c47b27cce846b.jpg"},{"id":84203682,"identity":"7777e3d0-5028-4fdd-9bdd-f322b65ccb6c","added_by":"auto","created_at":"2025-06-09 08:45:24","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":85879,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of reorientation performance between the Boundary and Non-boundary (NB) conditions, measured by angular difference error in recalling the Home direction. Each dot in the scatter plot represents the average value for an individual participant. White dots indicate group means, with error bars denoting standard deviations. Note that outliers were not excluded, as their removal would make the result more significant. ***: \u003cem\u003ep\u003c/em\u003e≤.001.\u003c/p\u003e","description":"","filename":"Figure3RE.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6491443/v1/6ef954d8186b83f97ee4dca7.jpeg"},{"id":84201990,"identity":"e5fde792-a983-4d97-b158-d581c2c3b532","added_by":"auto","created_at":"2025-06-09 08:37:24","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":242578,"visible":true,"origin":"","legend":"\u003cp\u003eN100 Component. (a) Grand average event-related potentials (ERPs) of the N100 for the Boundary and Non-boundary (NB) conditions. Due to individual variability in N100 waveforms, two peaks appear around 100 ms. The mean N100 amplitude was calculated within a time window (highlighted in yellow), defined as the mean ± standard deviation (96 ±18 ms) of the minimum peak latency across all participants. The vertical dashed line indicates the mean latency. (b) Distribution of minimum peak latency across all participants. Each black dot in the scatter plot represents an individual participant's minimum peak latency. White dots indicate group means, with error bars denoting standard deviations.\u003c/p\u003e","description":"","filename":"Figure4GrandERpeak.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6491443/v1/f32a388080c5f5f9a38d12bd.jpeg"},{"id":84201989,"identity":"556269f3-61dd-4070-a720-ff136666a043","added_by":"auto","created_at":"2025-06-09 08:37:24","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":80708,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of objective cognitive workload between the Boundary and Non-boundary (NB) conditions, measured by the mean amplitude of the N100 component. Each dot in the scatter plot represents the average value for an individual participant. White dots indicate group means, with error bars denoting standard deviations. ***: \u003cem\u003ep\u003c/em\u003e ≤.001.\u003c/p\u003e","description":"","filename":"Figure5meanAngle.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6491443/v1/e2a6972e550a14ad4a7395bb.jpeg"},{"id":84201992,"identity":"6f20f5eb-5916-4a15-9987-fa568711998b","added_by":"auto","created_at":"2025-06-09 08:37:25","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":81266,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of subjective workload between the Boundary and Non-boundary (NB) conditions, as measured by NASA-TLX scores. Each dot in the scatter plot represents the average value for an individual participant. White dots indicate group means, with error bars denoting standard deviations. ***: \u003cem\u003ep\u003c/em\u003e≤.001.\u003c/p\u003e","description":"","filename":"Figure6nasa.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6491443/v1/7e40505e5c8b826741908f85.jpeg"},{"id":102292362,"identity":"9fb42e67-2264-4183-a8ac-9b307a8ba8af","added_by":"auto","created_at":"2026-02-10 09:27:45","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1496022,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6491443/v1/97012c07-5c0b-4c03-a85c-37985cf96c70.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Virtual environmental boundaries reduce cognitive workload for reorientation during turn-by-turn navigation","fulltext":[{"header":"1.\tIntroduction","content":"\u003cp\u003eCurrent navigation assistants help users reach their destinations efficiently, addressing the challenge of traditional wayfinding, which often involves extensive trial and error. However, after following turn-by-turn instructions (such as turning left at the next intersection), GPS users can reach their destination but still feel disoriented, uncertain of where they are and which direction they are heading. This phenomenon highlights two distinct navigation strategies. The first, route following, requires no encoding of location or direction; instead, users rely on a fixed sequence of responses to specific stimuli, such as turn-by-turn instructions, to reach their destinations\u003csup\u003e{\u003c/sup\u003e\u003csup\u003e1}\u003c/sup\u003e.\u0026nbsp;The second strategy involves spatial knowledge acquisition, referred to as the reorientation strategy, supporting heading recovery and self-location estimation in the allocentric reference frame\u003csup\u003e{2,3,4}\u003c/sup\u003e. While route following reduces cognitive demands for reaching destinations, it does not support reorientation due to the absence of spatial information encoding, such as directions and locations. In contrast, the reorientation strategy, which involves the spatial knowledge acquisition, contributes to spatial memory and flexible navigation. For example, by encoding the relative location between an initial starting and a goal, the individual can find an alternative route to come back home when a familiar path is blocked\u003csup\u003e{2,3}\u003c/sup\u003e. Previous research found that users grow increasingly dependent on turn-by-turn navigation, weakening their spatial memory and independent navigation skills\u003csup\u003e{5,6,7}\u003c/sup\u003e. It raises a question: how do we balance the efficiency of reaching a destination and the effectiveness of spatial knowledge acquisition during GPS navigation?\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ePrevious research suggests that current navigation aids struggle to support both strategies simultaneously. Studies on navigation interfaces indicate a trade-off between these two strategies\u003csup\u003e\u0026nbsp;{8,9,10}\u003c/sup\u003e. A north-up map view, which provides an allocentric reference frame, enables users to quickly judge their location relative to landmarks but requires additional response time to follow turn-by-turn instructions. In contrast, a track-up map or real-street view aligns turn-by-turn route indications with users' egocentric reference frame, facilitating destination-reaching but leading to poor reorientation performance. Although users can mentally switch between egocentric and allocentric reference frames, misalignment between the two spatial reference frames increases response time and error rates of the mental rotation processing\u003csup\u003e{11,12}\u003c/sup\u003e. Furthermore, navigation aids divert attention from the physical environment, increasing the risk of attention distraction and reducing spatial knowledge acquisition\u003csup\u003e{6,7\u003c/sup\u003e\u003csup\u003e,13\u003c/sup\u003e\u003csup\u003e}\u003c/sup\u003e. These studies suggest that reorientation during turn-by-turn navigation imposes a high cognitive workload. When users prioritize reaching their destination via the shortest and most direct path, they tend to focus on turn-by-turn instructions for convenience, neglecting spatial information acquisition necessary for developing spatial ability.\u003c/p\u003e\n\u003cp\u003ePrevious research has attempted to encourage users to adopt spatial-knowledge-related strategies by presenting virtual cues that provide allocentric orientation in real environments. For example, compared to users who followed turn-by-turn instructions, those who navigated using auditory beacons from the direction of the destination demonstrated improved relative location estimation\u003csup\u003e{14}\u003c/sup\u003e. Similarly, integrating virtual global landmarks that provide stable orientation (such as distant towers and mountains) has been shown to enhance spatial memory compared to navigation guided solely by turn-by-turn instructions\u003csup\u003e\u0026nbsp;{15}\u003c/sup\u003e. However, in these studies, participants repeatedly encountered the same locations and routes, reinforcing their learning of paths and landmarks through repetition. This reinforcement learning process requires extensive environmental exploration, which contrasts with real-world GPS usage, where users typically minimize exploration, especially when navigating under time constraints, such as during emergency evacuations or urgent business travel in unfamiliar places. It remains unclear whether these methods are effective for reorientation in routes traveled only once, where users cannot rely on reinforcement learning of paths and landmarks. Moreover, previous studies have not demonstrated that virtual cues reduce the cognitive demands of spatial processing during navigation. We suggest that these studies may have underestimated users' preference for minimizing exploration and maximizing navigation efficiency. If virtual orientation cues do not reduce the cognitive demands of spatial knowledge strategies, users may prioritize turn-by-turn instructions. In other words, users are likely to minimize their effort to acquire spatial information if they can reach their destination without it.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; To address this issue, we argue that navigational aids should both enhance spatial knowledge acquisition and reduce the associated cognitive workload. In our previous research, we proposed visualizing virtual environmental boundaries as a global geometric reference that encloses the navigation area, providing intuitive support for direction and distance evaluation during turn-by-turn navigation\u003csup\u003e{16}\u003c/sup\u003e. Specifically, we introduced Augmented Reality (AR) City Walls, a design prototype that overlays encircling walls in the background of the user's field of view, positioned along the perimeter of a square navigation area defined by feasible routes between the starting point and the destination (see\u0026nbsp;Fig.\u0026nbsp;1). The design was inspired by geometric module theory, which suggests that animals possess an innate cognitive mechanism to process spatial information based on environmental geometry, such as walls of rooms and shapes of navigation area\u003csup\u003e{17,18,19}\u003c/sup\u003e. For example, rodents use specific sides or corners of a chamber's geometric boundary to locate the hidden food or judge their position relative to the boundary\u003csup\u003e{20,21,22}\u003c/sup\u003e. Human studies similarly support the role of boundaries in orientation\u003csup\u003e{20,23,24}\u003c/sup\u003e and relative location evaluation\u003csup\u003e{25,26}\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eAlthough our previous study had found that incorporating turn-by-turn indication with AR City Walls can improve reorientation performance without extensive exploration of environment\u003csup\u003e{16}\u003c/sup\u003e, it remains unclear whether environmental boundaries can reduce the cognitive workload associated with reorientation during turn-by-turn navigation. Since high-visibility geometric cues are rare in large-scale outdoor environments, no prior research has objectively measured the cognitive workload involved in using such cues within turn-by-turn navigation.\u003c/p\u003e\n\u003cp\u003eIn this study, we examine the effects of environmental boundaries (i.e., AR City Walls) on reorientation performance at the arrival of novel locations and the cognitive workload for using them during turn-by-turn navigation. We hypothesize that the virtual environmental boundaries can improve reorientation performance and reduce the cognitive workload when users follow a predefined route in unfamiliar environments, thereby facilitating efficient spatial knowledge acquisition while preserving the benefits of turn-by-turn navigation. We emphasize that both improved reorientation performance and reduced cognitive load are necessary conditions. If either condition is not met, environmental boundaries cannot efficiently improve spatial cognition during turn-by-turn navigation. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo test this hypothesis, we used mobile electroencephalography (EEG) to record brain activity while participants navigated simulated large-scale outdoor environments using virtual reality (VR) headsets. The VR headsets provided a realistic and expansive pedestrian perspective, allowing participants to control their movement direction through head rotations\u003csup\u003e{27,28}\u003c/sup\u003e. We recruited 42 participants, each experiencing two conditions (i.e., Non-boundary and Boundary). After excluding individuals affected by VR sickness and EEG signal artifacts, 35 participants remained for the final analysis. To minimize the influence of extensive exploration and ensure sufficient data sampling, each participant navigated both conditions twice in different environments.\u0026nbsp;Figure 2\u0026nbsp;shows one of the environments used in the experiment. The environment was visually occluded by dense forest, preventing visibility of other locations or distinctive landmarks. Each environment contained 25 specific location areas, with one centrally located 'Home' serving as both the initial starting point and the destination of the journey. A circular, winding main road connected these areas, ensuring that each location could be visited only once, except for Home. No identifying signs or landmarks were displayed at any location area, making each area visually indistinct from the surrounding road (pink and green areas surrounded by yellow circles in\u0026nbsp;Fig.\u0026nbsp;2\u0026nbsp;for illustration purposes only and were not visible in the VR environment). Participants were informed that 'Home' was their initial standing point, which was randomized for each environment to encourage self-guided spatial encoding. To induce disorientation, the roads between location areas included one or two U-turns. In both conditions, participants were instructed to face Home upon entering a new location area. The angular difference between their response and the actual direction of Home was recorded as a measure of reorientation performance. After responding, an arrow appeared to indicate the next travel direction, simulating turn-by-turn navigation. The arrow disappeared once participants exited the location area. In the Boundary condition, AR City Walls were displayed at two key points: before the start of travel, to aid initial orientation, and after the exit of location areas, to support reorientation. The walls disappeared after the left of the initial start point or the entry of the location areas. These designs limited continuous exposure to AR City Walls and required participants to rely on spatial knowledge acquired during navigation to recall Home's direction.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWe used the task-irrelevant auditory probe technique to assess objective cognitive workload for several reasons. This classical EEG paradigm estimates the attentional resources allocated to a task, particularly for complex and sustained activities such as driving\u003csup\u003e{29,30}\u003c/sup\u003e and video watching\u003csup\u003e{33}\u003c/sup\u003e. Compared to subjective self-reports\u003csup\u003e{29,33}\u003c/sup\u003e, such as the NASA-TLX questionnaire\u003csup\u003e{32}\u003c/sup\u003e, it provides a more accurate and objective assessment of the cognitive demands required for task completion without disrupting execution, as the task relies on non-auditory sensory modalities. Based on the principle that the available mental resources are limited at a given time\u003csup\u003e{33}\u003c/sup\u003e, participants passively listened to a series of auditory probes while performing target tasks. Event-related potentials (ERPs) such as N100 and P300 components were triggered by those task-irrelevant auditory probes, the amplitudes of which decrease as the cognitive workload of the target task increases\u003csup\u003e{29,33,34}\u003c/sup\u003e. In our experiment, auditory probes were presented continuously as background sounds. Only EEG epochs between the participant's departure from the previous location area and arrival at the next location area were selected for analysis, as this period corresponded to the display of the AR City Walls, the sole difference between the settings of Non-boundary and Boundary conditions. The mean N100 amplitude triggered by these probes at the FCz channel was calculated, as it is the most sensible component to assess objective cognitive workload according to previous research\u003csup\u003e{29,33}\u003c/sup\u003e. Additionally, participants completed the NASA-TLX questionnaire after navigation to provide subjective workload measures. Together, N100 mean amplitudes and NASA-TLX scores provided a comprehensive and comparable evaluation of cognitive workload.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn summary, we hypothesized that the presence of virtual environmental boundaries would enhance reorientation performance and reduce cognitive workload during turn-by-turn navigation.\u003c/p\u003e"},{"header":"2.\tResults","content":"\u003cp\u003eGiven our within-subject design (Non-boundary vs. Boundary), the two-tailed paired-sample t-test with a 95%\u0026nbsp;confidence interval was used for statistical analysis.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.1.\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eEnvironmental boundary improves reorientation performance\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eReorientation performance was assessed by calculating the average angular difference between the actual direction of Home and the direction participants faced when instructed to orient toward it. Results showed that participants were significantly more accurate in estimating their relative position to Home in the Boundary condition (M =\u0026nbsp;19.98,\u0026nbsp;SD =\u0026nbsp;10.68) compared to the Non-boundary condition (M =\u0026nbsp;57.96,\u0026nbsp;SD =\u0026nbsp;16.37),\u0026nbsp;t(34) = -15.00,\u0026nbsp;\u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026lt;\u0026nbsp;0.001,\u0026nbsp;CI[-43.13, -32.84]\u0026nbsp;(see\u0026nbsp;Fig. 3). These findings suggest that the presence of environmental boundaries significantly enhances reorientation performance during turn-by-turn navigation without requiring extensive environmental exploration.\u003c/p\u003e\n\u003cp\u003eNotably, since the AR City Walls were not displayed while participants navigated the location areas or made directional responses, the observed improvement in reorientation accuracy can be attributed to intermittent rather than continuous exposure to the boundaries.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.2.\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eEnvironmental boundary reduce the cognitive workload for reorientation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.2.1.\u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eObjective cognitive workload measured by N100 mean amplitude\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFigure 4\u0026nbsp;presents the grand average event-related potentials (ERPs) at electrode FCz for both conditions across all participants. The peak of the N100 component is visibly higher in the Non-boundary condition compared to the Boundary condition.\u003c/p\u003e\n\u003cp\u003eMean N100 amplitude was calculated within the 78–114 ms time window. Results showed that the Boundary condition (M = -2.20,\u0026nbsp;SD =\u0026nbsp;1.27) had a significantly lower mean amplitude than the Non-boundary condition (M = -1.86$,\u0026nbsp;SD =\u0026nbsp;1.15),\u0026nbsp;t(34) = -3.79,\u0026nbsp;\u003cem\u003ep\u003c/em\u003e =\u0026nbsp;0.001,\u0026nbsp;CI[-0.53, -0.16]\u0026nbsp;(see\u0026nbsp;Fig. 5). Since lower (i.e., more negative) N100 amplitudes are associated with reduced cognitive workload, this result indicates that the Boundary condition imposed less cognitive demand than the Non-boundary condition.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2.2.2.\u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eSubjective workload measured by NASA-TLX scores\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSubscale weightings were applied in NASA-TLX to compute overall workload scores. Since the only difference between the Boundary and Non-boundary conditions was the perception of AR City Walls, we interpret the overall NASA-TLX score as an indicator of cognitive demand. Higher scores reflect greater cognitive workload during the GPS navigation task. Results showed that the Boundary condition (M =\u0026nbsp;50.68,\u0026nbsp;SD =\u0026nbsp;14.79) was associated with significantly lower cognitive workload than the Non-boundary condition (M =\u0026nbsp;74.15,\u0026nbsp;SD =\u0026nbsp;8.21),\u0026nbsp;t(34) = -8.91,\u0026nbsp;\u003cem\u003ep\u003c/em\u003e \u0026lt;\u0026nbsp;0.001,\u0026nbsp;CI[-28.83, -18.12]\u0026nbsp;(see\u0026nbsp;Fig. 6).\u003c/p\u003e\n\u003cp\u003eThis subjective workload measure aligns with the objective cognitive workload indexed by the N100 mean amplitude. Together, these findings suggest that displaying environmental boundaries reduces the cognitive demands associated with reorientation during turn-by-turn navigation.\u003c/p\u003e"},{"header":"3.\tDiscussion","content":"\u003cp\u003eThis study aimed to investigate the effect of virtual environmental boundaries on the efficiency of spatial knowledge acquisition. We measured reorientation performance and assessed cognitive workload for spatial knowledge acquisition using EEG data during navigation, and a self-report questionnaire was administered afterward. In a within-subjects design, 42 participants navigated unfamiliar environments under both non-boundary and boundary conditions. To prevent extensive exploration of environments, each participant followed a predetermined route, ensuring that they were not exposed to the same locations or paths multiple times. We hypothesized that virtual environmental boundaries would enhance spatial information acquisition during turn-by-turn navigation if both criteria were met: (1) improved reorientation accuracy and (2) reduced cognitive workload during navigation.\u003c/p\u003e\n\u003cp\u003eOur results indicate that displaying virtual environmental boundaries enhances reorientation performance and reduces cognitive workload during turn-by-turn navigation. Participants demonstrated greater accuracy in estimating the relative location of Home after traveling with environmental boundaries for part of the route. This improvement in reorientation performance was accompanied by a decrease in cognitive workload, as reflected in the N100 mean amplitude and NASA-TLX scores. These findings suggest that virtual environmental boundaries can support efficient spatial information acquisition while preserving the benefits of turn-by-turn navigation in reaching a destination.\u003c/p\u003e\n\u003cp\u003eOur experimental setting was intentionally designed to avoid repeated exposure to the same locations or paths. As a result, the observed improvement in spatial knowledge acquisition can be primarily attributed to the presence of virtual environmental boundaries rather than interactions with other environmental cues. This implies that the environmental boundary can play a prominent role in spatial knowledge acquisition during large-scale outdoor navigation. The geometric module theory provides a plausible explanation for the role of virtual environmental boundaries in spatial processing: the global geometric layout of environmental boundaries offers intuitive direction and distance evaluation within a stable allocentric reference frame\u003csup\u003e{17,18,19}\u003c/sup\u003e. Unlike navigation aids based on map views\u003csup\u003e{8,9,11,12}\u003c/sup\u003e, which require substantial effort to align egocentric and allocentric reference frames, environmental boundaries may facilitate an automatic transformation between these spatial reference frames, thereby reducing the cognitive workload associated with reorientation.\u003c/p\u003e\n\u003cp\u003eUnlike previous research that assessed spatial knowledge after reinforcement learning of paths and landmarks\u003csup\u003e{14,15}\u003c/sup\u003e, our study focused on spatial knowledge acquisition during one-time navigation. This approach provides a more accurate evaluation of the efficiency of navigation aids in spatial information acquisition.\u003c/p\u003e\n\u003cp\u003eThis study highlights that users tend to maximize benefits while minimizing costs. When the primary goal of navigation is to reach a destination, if the cognitive workload required for spatial information acquisition is too high and spatial knowledge is not necessary for reaching the goal, users may choose to disregard spatial-knowledge-related navigation strategy. Based on this premise, this study leveraged EEG, specifically the N100 component, to provide an objective neural measure of cognitive workload during navigation. Compared to traditional self-report methods, EEG offers high temporal resolution and can capture subtle, real-time fluctuations in cognitive processing that may not be consciously perceived by participants. By integrating both subjective (NASA-TLX) and objective (N100 amplitude) data, our approach enhances the validity of the findings and demonstrates the value of neurophysiological methods in evaluating spatial cognition. This methodology offers a powerful tool for future research aiming to understand and improve the cognitive efficiency of navigation systems.\u003c/p\u003e\n\u003cp\u003eIn brief, we suggest that enhanced reorientation performance increases the benefits of navigation, while a reduced cognitive workload lowers the cost, allowing users to naturally utilize virtual environmental boundaries to improve their spatial knowledge while retaining the advantage of turn-by-turn instructions for route following.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHowever, like many previous studies, our research explicitly instructed participants to perform spatial cognition tasks (e.g., pointing tasks, relative location estimation), which may have encouraged active acquisition and processing of spatial information. Future studies should examine spatial processing efficiency in real-world field settings, where users engage in typical navigation behaviors using GPS systems without explicit instructions to acquire spatial knowledge. Additionally, we did not assess long-term spatial memory or cognitive map representation after complete learning. Future research should explore both the efficiency of spatial encoding and its long-term impact on cognitive map development. Furthermore, as our experiment simulated pedestrian navigation in a forested environment with sparse environmental cues and virtual reality constraints on participant movement, the impact of environmental boundaries should be evaluated in a broader range of real-world scenarios, such as driving and urban navigation, for practical applications.\u003c/p\u003e"},{"header":"4.\tMaterials and Methods","content":"\u003cp\u003e\u003cstrong\u003e4.1.\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eParticipants\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eForty-two individuals with normal or corrected-to-normal vision were recruited via a local online bulletin board. The sample included 20 women (mean age = 25.7 years, SD = 2.6, range = 21–32). All participants were right-handed and provided written informed consent before the study. Ethical approval was obtained from the local ethics committee (Internal Review Board, National Institute of Advanced Industrial Science and Technology, Japan; Protocol No. 2019-481). All experiments were conducted in accordance with relevant institutional guidelines and regulations. Due to simulator sickness and EEG data artifacts, seven participants were excluded, resulting in a final sample of 35. All participants followed the same experimental procedure.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.2.\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eEquipment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.2.1.\u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eVirtual\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003er\u003c/strong\u003e\u003cstrong\u003eeality\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHTC Vive Pro Eye head-mounted display (HTC Corporation, Taoyuan, Taiwan) with a 90 Hz refresh rate (Dual OLED 3.5\" diagonal screen, 1440 × 1600 pixels per eye, 120 Hz gaze data output frequency, 615 ppi, and 110°\u0026nbsp;nominal field of view) was used to present the virtual environment. The VR headset was equipped with a mobile EEG device. The EEG amplifier and data transmission unit were carried in a backpack (see\u0026nbsp;Fig 2a). For more details about EEG equipment, see the section about EEG recordings and analysis.\u003c/p\u003e\n\u003cp\u003eThe VR setup was placed in a 3.5 m × 3.5 m room with a ceiling height of 2.5 m, while the virtual environment spanned an area of 500 m × 500 m. To prevent movement obstruction, the VR headset cable was suspended from the center of the ceiling. A VR motion tracker monitored participants' head position and orientation. The movement of participants was a combination of the head rotation recorded by the VR movement tracker and trackpad on the left-hand controller. The positive and negative y-axes of the trackpad shifted forward and backward, respectively. The movement speed (1.2 ± 1.2 m/s) increased from the middle to both ends of the axis. The direction of movement was aligned with the heading in a 2-dimension plane parallel to the transverse section of the individual head. The back-and-forth pan, controlled by the left trackpad and multiplied by the head rotation, resulted in smooth movement. This movement prevents the occurrence of simulation sickness induced by bouncing up and down during actual walking. The trigger button on the right-hand controller was used to interact with the visual instruction window, a 2-dimensional panel right in front of the view of participants.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.2.2.\u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eMap layouts\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFour maps for the navigation experiment were built using the Unity3D game engine (Unity Technologies, San Francisco, California, USA): one for practice and the other three for the formal experiment (for an example, see\u0026nbsp;Fig. 2). The navigation area of each map was 500 m × 500 m. A winding, closed-loop road passing through the forest formed the main road of each map.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe main road in the practice map passed through 19 location areas. In the formal test, each of the three main roads traversed 25 location areas, including a central 'Home' location that served as both the starting point and the final destination. Thirteen of these location areas were configured as nested hexagons on each map (rendered in green in\u0026nbsp;Fig. 2b). These two hexagons had the same center, which was the home. The side length (240 m) of the first hexagon was twice that of the second. The sequence of roads connecting these 13 fixed locations varied across maps. The remaining location areas (6 in the practice map and 12 in each formal test map, rendered in pink in\u0026nbsp;Fig.\u0026nbsp;2b) were positioned after connecting the 13 fixed points with a circular road. These additional location areas were placed based on the condition that one or two turns must occur between any two consecutively visited areas on the main route. The main roads lacked clear geometric information, such as right angles and straight lines. The roads were intentionally designed without clear geometric features such as right angles or straight paths, to enhance disorientation between successive locations.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eParticipants' views were obstructed by dense forest, preventing them from seeing other location areas. Apart from trees and vegetation, no significant landmarks were present. Moreover, no identifying signs or markers were displayed at any of the location areas, making each one visually indistinguishable from other parts of the road. While each location area was defined by a 12-meter radius circle, this circle was not visible in the VR environment.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.2.3.\u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eNavigation aids\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants navigated the virtual environment using different navigation aids depending on the condition: Boundary or Non-boundary.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTurn-by-turn indication (used in both conditions)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn both conditions, participants received turn-by-turn navigation assistance. A virtual arrow appeared at each location area, indicating the next forward direction. Participants were instructed to follow the arrow to reach the next location. Once they passed over the location area, the arrow disappeared. The arrow was not displayed during travel along the road between two location areas.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAR City Walls (used only in the boundary condition)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe AR City Walls served as a virtual environmental boundary and acted as a high-visibility geometric cue providing an allocentric reference frame for orientation and distance estimation. This boundary consisted of four virtual city walls placed in the cardinal directions (i.e., North, South, East, and West) enclosing the square-shaped navigation environment.\u003c/p\u003e\n\u003cp\u003eThe distinct visual features of each wall enabled participants to identify the cardinal directions. The north wall featured a blue base with a central spire-topped building, while the south wall had a similar structure but was gray. The east and west walls were distinguished by white bases and central towers, red on the east and blue on the west (see\u0026nbsp;Fig. 2b).\u003c/p\u003e\n\u003cp\u003eWhen participants stood at the center of the environment facing north, the northern city wall with the blue spire was directly in front of them. As they moved northward, the northern wall appeared progressively larger and more prominent in their field of view (see\u0026nbsp;Fig. 1c), reinforcing their perception of direction and distance within the environment.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.2.4.\u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eTask-irrelevant auditory probe\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTask-irrelevant auditory probes were displayed as the background sounds in each run and programmed using MATLAB (MathWorks, Inc.). Similar to previous studies\u003csup\u003e{29,30,33}\u003c/sup\u003e, the probe sequence consisted of 12 pure tones ranging from 500 to 1600 Hz at intervals of 100 Hz. The duration of each probe was 50 ms, including rise and fall times of 10 ms. All tones were presented randomly with equal probability under the precondition that probes with the same frequency did not appear successively. The stimulus interval was randomized in the range of 400–800 ms (mean: 600 ms). The probe sequence was provided at approximately 75 dB/SPL using earphones on the VR headset. Participants were asked to neglect the background sounds, and none reported discomfort due to the probe presentation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.3.\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eExperimental procedure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.3.1.\u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eNavigation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDuring the experiment, participants were never exposed to a global map or an aerial view of the navigation area. A run was defined as a full traversal of the main road on a given map, returning to the starting point. Participants completed two runs during the practice session using the same map, and four runs during the formal test using three different maps. Each run followed an identical procedure; however, the practice session was designed to be less challenging than the formal test. Therefore, we describe the formal test procedure first.\u003c/p\u003e\n\u003cp\u003eIn the formal test, each condition (Boundary and Non-boundary) consisted of two runs, resulting in a total of four runs per participant. A fully randomized design ensured that no condition was paired with the same map layout more than once, and that the same condition and layout were not repeated consecutively. The condition–layout combinations and run order were counterbalanced across participants to control for order effects. Each run lasted approximately 20 minutes. After completing each navigation run, participants removed the VR headset, filled out the NASA-TLX questionnaire on a laptop using mouse input, and then took a short break.\u003c/p\u003e\n\u003cp\u003eAt the beginning of each run, participants started at Home, the central point of the navigation area. While at Home, participants were stationary but could freely rotate their heads to encode their initial position. In both conditions, the first directional arrow indicating the starting direction was visible. In the Boundary condition, participants could also perceive the AR City Walls at this stage. Participants were informed that the Home location was randomized for each run, and none of them knew its position in advance. Once participants felt confident, they had encoded the Home location, they pressed the right-hand controller trigger to begin the navigation. Upon confirmation, the white directional arrow turned blue in both conditions, and the AR City Walls disappeared in the Boundary condition.\u003c/p\u003e\n\u003cp\u003eEach time participants navigated from one location area to the next, a trial was completed. Accordingly, each formal test run included 24 trials. Each trial followed the same procedure. After leaving the current location area, the directional arrow at that location disappeared in both conditions. In the Boundary condition, the AR City Walls were re-displayed only after the participant exited the current location area, while no additional navigation aids were provided in the Non-boundary condition. Upon entering the next location area, participants were again stationary but allowed to rotate their heads to observe the surroundings. At this point, the AR City Walls (if visible) automatically disappeared. Then, a visual instruction window appeared, prompting participants to face toward the Home location, which was not visible. Participants indicated their estimated direction to Home by physically turning to face it and pressing the right trigger to confirm their response. No feedback was given about the accuracy of their estimation. Reorientation performance was assessed by calculating the angular difference between the participant's facing direction and the actual direction to Home. Next, a visual instruction reminded participants to observe the directional arrow, which appeared after they confirmed their Home direction. After pressing the right-hand trigger again, the instruction window disappeared, and the arrow turned blue, signaling that they could proceed to the next location. If participants moved in the wrong direction, a warning window appeared in front of them, guiding them back to the previous location area. This process was repeated until they completed the full loop and returned to Home.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe road length in the practice sessions was shorter than in the formal experiment, with 19 location areas in the practice map compared to 25 in each of the formal test layouts. During the practice runs, navigation aids, including the directional arrows and AR City Walls, were continuously displayed from start to finish in both runs. This approach was intended to help participants become familiar with the navigation aids and task procedures. Participants were allowed to perceive the AR City Walls when estimating the Home direction during practice. They could also stop the practice session at any time, and completion of the NASA-TLX questionnaire was not required. Data collected during the practice session were excluded from the final analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.3.2.\u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eEEG recordings and analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEEG activity was recorded using the BrainAmp Standard amplifier, the BrainVision MOVE 32-channel wireless transmission system, and the actiCAP slim active EEG electrode system (Brain Products GmbH, Germany). Data analysis was performed using the EEGLAB toolbox\u003csup\u003e{35}\u003c/sup\u003e in MATLAB. During navigation, the transmission system was secured in a backpack worn by participants.\u003c/p\u003e\n\u003cp\u003eThe EEG system used 28 electrode sites (Fp1, Fp2, F7, F3, Fz, F4, F8, FCz, T7, C3, Cz, C4, T8, TP7, CP3, CPz, CP4, TP8, P7, P3, Pz, P4, P8, POz, O1, Oz, O2, and M2 in the extended 10–20 System), with M1 and AFz as the reference and ground electrodes, respectively. Horizontal and vertical electrooculograms (EOG) were recorded using electrodes placed on the outer left and right canthi and above and below the right eye. The impedance of all electrodes was maintained below 10 kΩ. EEG signals were digitized at a sampling rate of 1000 Hz, and the time constant was set to 10 s.\u003c/p\u003e\n\u003cp\u003ePreprocessing was conducted to clean the raw EEG data and increase the signal-to-noise ratio of the movable EEG datasets. The EEG signals were re-referenced to the averaged EEG signals from the linked mastoids (M1 and M2) and band-pass filtered at 0.1–30 Hz. Artifacts derived from the EOG and body movements were removed from the data using independent component analysis\u003csup\u003e{35}\u003c/sup\u003e. To compute ERPs elicited by probe stimuli during the reorientation stage, averaged epochs were extracted from the EEG segmentation of the specific stage with a data window of 500 ms, including a 100 ms pre-stimulus baseline period.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ePrior to ERP averaging, epochs in which the EEG signal variation exceeded ± 80 μV were removed. In each condition, epochs from the two runs were combined to ensure adequate data sampling after artifact rejection (the mean remaining epoch numbers for the Boundary and Non-boundary conditions were 630 and 650, respectively). As reported in previous research\u003csup\u003e{29,33}\u003c/sup\u003e, the mean N100 amplitude at FCz is the most sensitive indicator of cognitive workload compared to that at the other electrode sites. Therefore, the average ERPs at the FCz were calculated. The grand average ERP at FCz across 35 participants showed two negative peaks around 100 ms, likely due to variability in individual ERP latencies (see\u0026nbsp;Fig. 4). However, an examination of each participant's ERP in each condition revealed only one maximum negative peak. To determine the appropriate time window for calculating the mean amplitude of N100, we first computed the mean (96 ms) and standard deviation (18 ms) of the N100 peak latency. Based on this, the mean amplitude of N100 was calculated as the mean amplitude in the time window of 78–114 ms. The N100 mean amplitude assessed the objective cognitive workload during navigation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4.4.\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eStatistical analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eStatistical analysis was conducted using SPSS 23 (IBM, Armonk, NY, USA). A two-tailed paired-sample t-test with a 95%\u0026nbsp;confidence interval was used to compare conditions. Three key measures were analyzed:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eReorientation performance\u003c/strong\u003e, assessed by the angular difference between the actual direction of Home and the direction participants faced when instructed to orient toward Home. The average reorientation performance across 48 trials (combining the first and second periods) was calculated.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eObjective cognitive workload\u003c/strong\u003e, measured by the mean amplitude of the N100 component. For each participant in each condition, the N100 mean amplitude was calculated within a time window of 78–114 ms.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSubjective cognitive workload\u003c/strong\u003e, assessed using NASA-TLX scores. In each run, NASA-TLX scores were computed as a weighted average of six subscales: mental demand, physical demand, temporal demand, performance, effort, and frustration. The average overall NASA-TLX score across the first and second periods was calculated for each condition. Since the perception of AR City Walls was the only difference between the Boundary and Non-boundary conditions, we suggest that the overall NASA-TLX score serves as a measure of cognitive demand.\u003c/p\u003e\n\u003cp\u003eThe Shapiro-Wilk test was used to assess the normality of all data samples. The results indicated that subjective and objective workload data were normally distributed. Although reorientation performance data did not follow a normal distribution, outliers were not removed to avoid artificially inflating the significance of the results (see Fig. 3).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank Ms. Shinobu Yasumuro at the National Institute of Advanced Industrial Science and Technology and Ms. Weiran Cui at the University of Tsukuba for their support in coordinating the experiment. This study was supported in part by the Japan Society for the Promotion of Science (JSPS) KAKENHI [grant numbers 21H03787 and 23K21851] and JST SPRING [grant number JPMJSP2124].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eXY.Z and S.I conceptualized the study and conducted data acquisition. XY.Z developed the methodology, performed data analysis, wrote the main manuscript, and prepared the figures. Both XY.Z and S.I reviewed and edited the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdditional information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003cstrong\u003e：\u003c/strong\u003e The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eCook, D. \u0026amp; Kesner, R. P. Caudate nucleus and memory for egocentric localization. \u003cem\u003eBehav. Neural Biol\u003c/em\u003e.\u003cstrong\u003e49\u003c/strong\u003e, 332-343, DOI: 10.1016/S0163-1047(88)90338- X (1988).\u003c/li\u003e\n\u003cli\u003eTolman, E. C. Cognitive maps in rats and men. \u003cem\u003ePsychol. Rev\u003c/em\u003e.\u003cstrong\u003e55\u003c/strong\u003e, 189\u0026OElig;208, DOI:10.1037/h0061626 (1948).\u003c/li\u003e\n\u003cli\u003eEpstein, R. A., Patai, E. Z., Julian, J. B. \u0026amp; Spiers, H. J. The cognitive map in humans: spatial navigation and beyond. \u003cem\u003eNat.\u003c/em\u003e\u003cem\u003e Neurosci\u003c/em\u003e.\u003cstrong\u003e20\u003c/strong\u003e, 1504-1513, DOI:10.1038/nn.4656 (2017).\u003c/li\u003e\n\u003cli\u003eJulian, J. B., Keinath, A. T., Marchette, S. A. \u0026amp; Epstein, R. A. The neurocognitive basis of spatial reorientation. \u003cem\u003eCurr.Biol\u003c/em\u003e. \u003cstrong\u003e28\u003c/strong\u003e, R1059-1073, DOI:10.1016/j.cub.2018.04.057 (2018).\u003c/li\u003e\n\u003cli\u003eDahmani, L. \u0026amp; Bohbot, V. D. Habitual use of gps negatively impacts spatial memory during self-guided navigation. \u003cem\u003eSci.Reports\u003c/em\u003e\u003cem\u003e \u003c/em\u003e\u003cstrong\u003e10\u003c/strong\u003e, 6310, DOI:10.1038/s41598- 020- 62877- 0 (2020).\u003c/li\u003e\n\u003cli\u003eGardony, A. L., Bruny\u0026eacute;, T. T., Mahoney, C. R. \u0026amp; Taylor, H. A. How navigational aids impair spatial memory: Evidence for divided attention. \u003cem\u003eSpatial Cognition. \u0026amp; Comput\u003c/em\u003e.\u003cstrong\u003e13\u003c/strong\u003e, 319-350, DOI:10.1080/13875868.2013.792821 (2013).\u003c/li\u003e\n\u003cli\u003eYing, Q., Dong, W. \u0026amp; Fabrikant, S. I. How do in-car navigation aids impair expert navigators spatial learning ability? \u003cem\u003eAnnals Am. Assoc. Geogr\u003c/em\u003e.\u003cstrong\u003e114\u003c/strong\u003e, 1483-1504, DOI:10.1080/24694452.2024.2356858 (2024).\u003c/li\u003e\n\u003cli\u003eM\u0026uuml;nzer, S., Zimmer, H. D. \u0026amp; Baus, J. Navigation assistance: a trade-off between wayfinding support and configural learning support. J. \u003cem\u003eexperimental psychology: applied\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 18-37, DOI:10.1037/a0026553 (2012).\u003c/li\u003e\n\u003cli\u003eM\u0026uuml;nzer, S., L\u0026ouml;rch, L. \u0026amp; Frankenstein, J. Wayfinding and acquisition of spatial knowledge with navigation assistance. \u003cem\u003eJ.\u003c/em\u003e\u003cem\u003e Exp. Psychol. Appl\u003c/em\u003e. \u003cstrong\u003e26\u003c/strong\u003e, 73-88, DOI:10.1037/xap0000237 (2020).\u003c/li\u003e\n\u003cli\u003eQiu, X., Yang, Z., Yang, J., Wang, Q. \u0026amp; Wang, D. Impact of ar navigation display methods on wayfinding performance and spatial knowledge acquisition. \u003cem\u003eInt. J. Human Computer Interact\u003c/em\u003e. \u003cstrong\u003e40\u003c/strong\u003e, 2676-2696, DOI:10.1080/10447318.2023.2169524 (2024).\u003c/li\u003e\n\u003cli\u003eAretz, A. J. \u0026amp; and, C. D. W. The mental rotation of map displays. \u003cem\u003eHum. Perform\u003c/em\u003e. \u003cstrong\u003e5\u003c/strong\u003e, 303-328, DOI:10.1207/s15327043hup0504_3 (1992).\u003c/li\u003e\n\u003cli\u003eShepard, R. N. \u0026amp; Hurwitz, S. Upward direction, mental rotation, and discrimination of left and right turns in maps. \u003cem\u003eCognition\u003c/em\u003e\u003cem\u003e \u003c/em\u003e\u003cstrong\u003e18\u003c/strong\u003e, 161-193, DOI:10.1016/0010- 0277(84)90024- 6 (1984).\u003c/li\u003e\n\u003cli\u003eIshikawa, T., Fujiwara, H., Imai, O. \u0026amp; Okabe, A. Wayfinding with a gaps-based mobile navigation system: A comparison with maps and direct experience. \u003cem\u003eJ. \u003c/em\u003e\u003cem\u003eEnviron. Psychol\u003c/em\u003e. \u003cstrong\u003e28\u003c/strong\u003e, 74-82, DOI:10.1016/j.jenvp.2007.09.002 (2008).\u003c/li\u003e\n\u003cli\u003eClemenson, G. D., Maselli, A., Fiannaca, A. J., Miller, A. \u0026amp; Gonzalez-Franco, M. Rethinking gps navigation: creating cognitive maps through auditory clues. \u003cem\u003eSci. Reports\u003c/em\u003e\u003cem\u003e \u003c/em\u003e\u003cstrong\u003e11\u003c/strong\u003e, 7764, DOI:10.1038/s41598- 021- 87148- 4 (2021).\u003c/li\u003e\n\u003cli\u003eLiu, J., Singh, A. K., Wunderlich, A., Gramann, K. \u0026amp; Lin, C.-T. Redesigning navigational aids using virtual global landmarks to improve spatial knowledge retrieval. \u003cem\u003enpj Sci. Learn\u003c/em\u003e. \u003cstrong\u003e7\u003c/strong\u003e, 17, DOI:10.1038/s41539- 022- 00132- z (2022).\u003c/li\u003e\n\u003cli\u003eZhang, X. \u0026amp; Iwaki, S. Self-location and reorientation of individuals without reading maps: Increased spatial memory during gps navigation using ar city walls. In Wei, J. \u0026amp; Margetis, G. (eds.) \u003cem\u003eHuman-Centered Design, Operation and Evaluation of Mobile Communications\u003c/em\u003e, 254-267, DOI:10.1007/978- 3- 031- 60458- 4_17 (Springer Nature Switzerland, Cham, 2024).\u003c/li\u003e\n\u003cli\u003eCheng, K. A purely geometric module in the rat\u0026apos;s spatial representation. \u003cem\u003eCognition\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 149-178, DOI:10.1016/0010- 0277(86)90041- 7 (1986).\u003c/li\u003e\n\u003cli\u003eTommasi, L., Chiandetti, C., Pecchia, T., Sovrano, V. A. \u0026amp; Vallortigara, G. From natural geometry to spatial cognition. \u003cem\u003eNeurosci. \u0026amp; Biobehav. Rev.\u003c/em\u003e\u003cem\u003e \u003c/em\u003e\u003cstrong\u003e36\u003c/strong\u003e, 799-824, DOI:10.1016/j.neubiorev.2011.12.007 (2012).\u003c/li\u003e\n\u003cli\u003eCheng, K. \u0026amp; Newcombe, N. S. Is there a geometric module for spatial orientation? squaring theory and evidence. \u003cem\u003ePsychon.\u003c/em\u003e\u003cem\u003e Bull. \u0026amp; Rev\u003c/em\u003e.\u003cstrong\u003e12\u003c/strong\u003e, 1-23, DOI:10.3758/BF03196346 (2005).\u003c/li\u003e\n\u003cli\u003eKeinath, A. T., Julian, J. B., Epstein, R. A. \u0026amp; Muzzio, I. A. Environmental geometry aligns the hippocampal map during spatial reorientation. \u003cem\u003eCurr. Biol.\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 309-317, DOI:10.1016/j.cub.2016.11.046 (2017).\u003c/li\u003e\n\u003cli\u003eWeiss, S.et al. Consistency of spatial representations in rat entorhinal cortex predicts performance in a reorientation task. \u003cem\u003eCurr. Biol\u003c/em\u003e. \u003cstrong\u003e27\u003c/strong\u003e, 3658-3665, DOI:10.1016/j.cub.2017.10.015 (2017).\u003c/li\u003e\n\u003cli\u003eLever, C., Burton, S., Jeewajee, A., O\u0026apos;Keefe, J. \u0026amp; Burgess, N. Boundary vector cells in the subiculum of the hippocampal formation. \u003cem\u003eJ. Neurosci.\u003c/em\u003e\u003cem\u003e \u003c/em\u003e\u003cstrong\u003e29\u003c/strong\u003e, 9771-9777, DOI:10.1523/JNEUROSCI.1319- 09.2009 (2009).\u003c/li\u003e\n\u003cli\u003eShine, J. P., Vald\u0026eacute;s-Herrera, J. P., Hegarty, M. \u0026amp; Wolbers, T. The human retrosplenial cortex and thalamus code head direction in a global reference frame. \u003cem\u003eJ. Neurosci.\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 6371-6381, DOI:10.1523/JNEUROSCI.1268- 15.2016 (2016).\u003c/li\u003e\n\u003cli\u003eMarchette, S. A., Vass, L. K., Ryan, J. \u0026amp; Epstein, R. A. Anchoring the neural compass: coding of local spatial reference frames in human medial parietal lobe. \u003cem\u003eNat. Neurosci.\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 1598-1606, DOI:10.1038/nn.3834 (2014).\u003c/li\u003e\n\u003cli\u003eBellmund, J. L. S.et al. Deforming the metric of cognitive maps distorts memory. \u003cem\u003eNat. Hum. Behav\u003c/em\u003e. \u003cstrong\u003e4\u003c/strong\u003e, 177-188, DOI:10.1038/s41562- 019- 0767- 3 (2020).\u003c/li\u003e\n\u003cli\u003eKeinath, A. T., Rechnitz, O., Balasubramanian, V. \u0026amp; Epstein, R. A. Environmental deformations dynamically shift human spatial memory. \u003cem\u003eHippocampus\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 89-101, DOI:10.1002/hipo.23265 (2021).\u003c/li\u003e\n\u003cli\u003eTaube, J. S., Valerio, S. \u0026amp; Yoder, R. M. Is navigation in virtual reality with fmri really navigation? \u003cem\u003eJ. Cogn. Neurosci\u003c/em\u003e. \u003cstrong\u003e25\u003c/strong\u003e,1008-1019, DOI:10.1162/jocn_a_00386 (2013).\u003c/li\u003e\n\u003cli\u003eMakeig, S., Gramann, K., Jung, T.-P., Sejnowski, T. J. \u0026amp; Poizner, H. Linking brain, mind and behavior. \u003cem\u003eInt. J . Psy-chophysiol\u003c/em\u003e. \u003cstrong\u003e73\u003c/strong\u003e, 95-100, DOI:10.1016/j.ijpsycho.2008.11.008 (2009). Neural Processes in Clinical Psychophysiology.\u003c/li\u003e\n\u003cli\u003eSugimoto, F.et al. Effects of one-pedal automobile operation on the driver\u0026apos;s emotional state and cognitive workload. \u003cem\u003eAppl. Ergonomics\u003c/em\u003e \u003cstrong\u003e88\u003c/strong\u003e, 103179, DOI:10.1016/j.apergo.2020.103179 (2020).\u003c/li\u003e\n\u003cli\u003eTakeda, Y., Inoue, K., Kimura, M., Sato, T. \u0026amp; Nagai, C. Electrophysiological assessment of driving pleasure and difficulty using a task-irrelevant probe technique. \u003cem\u003eBiol. Psychol\u003c/em\u003e. \u003cstrong\u003e120\u003c/strong\u003e, 137-141, DOI:10.1016/j.biopsycho.2016.09.009 (2016).\u003c/li\u003e\n\u003cli\u003eTakeda, Y. \u0026amp; Kimura, M. The auditory n1 amplitude for task-irrelevant probes reflects visual interest. Int. \u003cem\u003eJ. Psychophysiol.\u003c/em\u003e\u003cem\u003e \u003c/em\u003e\u003cstrong\u003e94\u003c/strong\u003e, 35-41, DOI:10.1016/j.ijpsycho.2014.07.007 (2014).\u003c/li\u003e\n\u003cli\u003eHart, S. G. \u0026amp; Staveland, L. E. Development of nasa-tlx (task load index): Results of empirical and theoretical research.In Hancock, P. A. \u0026amp; Meshkati, N. (eds.) \u003cem\u003eHuman Mental Workload\u003c/em\u003e, vol. 52 of \u003cem\u003eAdvances in Psychology\u003c/em\u003e, 139-183, DOI:10.1016/S0166- 4115(08)62386- 9 (North-Holland, 1988).\u003c/li\u003e\n\u003cli\u003eKahneman, D. Attention and effort, vol. 1063 (Prentice-Hall, 1973).\u003c/li\u003e\n\u003cli\u003eKramer, A. F., Trejo, L. J. \u0026amp; Humphrey, D. Assessment of mental workload with task-irrelevant auditory probes. \u003cem\u003eBiol.Psychol.\u003c/em\u003e \u003cstrong\u003e40\u003c/strong\u003e, 83-100, DOI:10.1016/0301- 0511(95)05108- 2 (1995). EEG in Basic and Applied Settings.\u003c/li\u003e\n\u003cli\u003eDelorme, A. \u0026amp; Makeig, S. Eeglab: an open source toolbox for analysis of single-trial eeg dynamics including independent component analysis. \u003cem\u003eJ. Neurosci. Methods\u003c/em\u003e \u003cstrong\u003e134\u003c/strong\u003e, 9-21, DOI:10.1016/j.jneumeth.2003.10.009 (2004).\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":"spatial cognition, navigation, navigation aids, cognitive workload, virtual reality, EEG","lastPublishedDoi":"10.21203/rs.3.rs-6491443/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6491443/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Current GPS navigation tools primarily help users reach their destinations through turn-by-turn instructions but offer limited support for reorientation, the ability to maintain a sense of direction and self-positioning. Because reorientation during turn-by-turn navigation demands a high cognitive workload, users prioritizing efficiency and safety tend to focus on following instructions rather than encoding their spatial bearings. To address this issue, we proposed visualizing virtual environmental boundaries, such as Augmented Reality (AR) City Walls in the background of the field of view, to serve as a global geometric reference surrounding the navigation area for intuitive direction and distance evaluation during turn-by-turn navigation. Using mobile electroencephalography (EEG), we assessed the cognitive workload of 35 participants as they navigated with virtual reality (VR) headsets. The results indicate that, compared to conventional turn-by-turn navigation using route indications only, displaying environmental boundaries enhances reorientation accuracy while reducing cognitive workload. These findings suggest a potential opportunity for GPS navigation to both the efficiency of reaching a destination and the effectiveness of spatial knowledge acquisition.","manuscriptTitle":"Virtual environmental boundaries reduce cognitive workload for reorientation during turn-by-turn navigation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-09 08:37:20","doi":"10.21203/rs.3.rs-6491443/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":"d050e69f-2946-448a-a07d-f9f92fcee427","owner":[],"postedDate":"June 9th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":49494195,"name":"Biological sciences/Neuroscience/Cognitive neuroscience"},{"id":49494196,"name":"Biological sciences/Neuroscience/Learning and memory/Spatial memory"},{"id":49494197,"name":"Biological sciences/Psychology/Human behaviour"},{"id":49494198,"name":"Physical sciences/Engineering"}],"tags":[],"updatedAt":"2026-02-10T09:26:28+00:00","versionOfRecord":[],"versionCreatedAt":"2025-06-09 08:37:20","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6491443","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6491443","identity":"rs-6491443","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}