Embodied Intelligence Enables Agentic Exploration in Microscopy

Embodied Intelligence Enables Agentic Exploration in Microscopy | 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 Embodied Intelligence Enables Agentic Exploration in Microscopy Huijun Gao, Gang Huang, Zhengyang Zhang, Songlin Zhuang, Xinghu Yu, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8617009/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Microscopy has fundamentally transformed our understanding of life; however, despite relentless technological iteration, its underlying operational paradigm remains strikingly static. Current instruments persist as passive data collectors, rigidly executing predefined commands while remaining blind to the stochastic and dynamic nature of biological systems. This reliance on passive observation creates a semantic gap between high-level scientific inquiry and physical execution, limiting the generalizability in adapting to uncertain tasks or dynamic biological events. To bridge this gap, we present an Embodied Intelligence Microscope System (EIMS) that reconceptualizes the microscope as an autonomous robot capable of active exploration. By physically grounding the reasoning capabilities of intelligent system, EIMS closes the loop between scientific intent, physical perception, and action, enabling a transition from hypothesis-driven verification to discovery-driven explore. The system achieves zero-shot generalization across complex, multi-step tasks and can autonomously formulate experimental protocols, navigate unknown spatiotemporal environments to locate regions of interest, and dynamically optimizes imaging parameters in real-time to resolve ambiguity. This work establishes a new paradigm for self-driving instruments, providing a generalizable framework where embodied intelligence acts not merely as a tool, but as an active partner in autonomous scientific discovery. Biological sciences/Biological techniques/Software Biological sciences/Biological techniques/Microscopy/Confocal microscopy Full Text Additional Declarations There is NO Competing Interest. Supplementary Files nmsupplementaryinformation.pdf Supplementary information SupplementaryVideo1.mp4 Overall introduction and multi-step experiment with fluorescence slice data acquisition and analysis. SupplementaryVideo2.mp4 Error correction experiment with fluorescence slice data acquisition and analysis. SupplementaryVideo3.mp4 Cross-platform generalization experiments. SupplementaryVideo4.mp4 Comparison with alternative embodied AI architectures. Cite Share Download PDF Status: Under Review 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-8617009","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":588395983,"identity":"f0222706-940f-4411-ade6-043210eb573a","order_by":0,"name":"Huijun Gao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvElEQVRIiWNgGAWjYBACA4YDQLLiAA+II0GCljMHeHhI0AIEjG0HGIjXYs54/OKHj/PuyNgzMB+8zcNgl0dQi2XDmWLJmdueAR3GlmzNw5BcTNhhB84kSPNuOwzUwmMmzcNwILGBCC3Jv3nngLTwfyNWy/Fj0rwNYFvYiLaFzXLGMaCWw2zGlnMMkonQcuP44xsfag7bs7c3P7zxpsKOsBYGiTMGEAYz2ASC6oGAv/0BMcpGwSgYBaNgJAMA2OM89DJFrUcAAAAASUVORK5CYII=","orcid":"","institution":"Research Institute of Intelligent Control and Systems, Harbin Institute of Technology, Harbin, China","correspondingAuthor":true,"prefix":"","firstName":"Huijun","middleName":"","lastName":"Gao","suffix":""},{"id":588395984,"identity":"427e80cf-86fb-4ab9-a08f-cd52c708cca3","order_by":1,"name":"Gang Huang","email":"","orcid":"","institution":"Research Institute of Intelligent Control and Systems, Harbin Institute of Technology, Harbin, China","correspondingAuthor":false,"prefix":"","firstName":"Gang","middleName":"","lastName":"Huang","suffix":""},{"id":588395985,"identity":"962df7f0-8b4d-4dfa-8c2d-d75518b2858f","order_by":2,"name":"Zhengyang Zhang","email":"","orcid":"","institution":"Research Institute of Intelligent Control and Systems, Harbin Institute of Technology, Harbin, China","correspondingAuthor":false,"prefix":"","firstName":"Zhengyang","middleName":"","lastName":"Zhang","suffix":""},{"id":588395986,"identity":"0958c2bb-2bae-4c07-afd1-21bdfd2c22bd","order_by":3,"name":"Songlin Zhuang","email":"","orcid":"","institution":"Yongjiang Laboratory, Ningbo, China","correspondingAuthor":false,"prefix":"","firstName":"Songlin","middleName":"","lastName":"Zhuang","suffix":""},{"id":588395987,"identity":"f8d6cb52-bb9f-4b98-8d84-c7b7c371f233","order_by":4,"name":"Xinghu Yu","email":"","orcid":"","institution":"Ningbo Institute of Intelligent Equipment Technology Company, Ltd., Ningbo, China","correspondingAuthor":false,"prefix":"","firstName":"Xinghu","middleName":"","lastName":"Yu","suffix":""},{"id":588395988,"identity":"c06567fd-71eb-4611-bdaf-cde27f22ed01","order_by":5,"name":"Yang Wu","email":"","orcid":"","institution":"Research Institute of Intelligent Control and Systems, Harbin Institute of Technology, Harbin, China","correspondingAuthor":false,"prefix":"","firstName":"Yang","middleName":"","lastName":"Wu","suffix":""},{"id":588395989,"identity":"e0b29dbb-d920-4271-aca7-52005444c4fe","order_by":6,"name":"Zhan Li","email":"","orcid":"","institution":"Research Institute of Intelligent Control and Systems, Harbin Institute of Technology, Harbin, China","correspondingAuthor":false,"prefix":"","firstName":"Zhan","middleName":"","lastName":"Li","suffix":""},{"id":588395990,"identity":"9de7c6a0-6f11-47af-83b3-51d92e33b000","order_by":7,"name":"Zhitai Liu","email":"","orcid":"","institution":"Research Institute of Intelligent Control and Systems, Harbin Institute of Technology, Harbin, China","correspondingAuthor":false,"prefix":"","firstName":"Zhitai","middleName":"","lastName":"Liu","suffix":""},{"id":588395991,"identity":"910c9253-8cfb-4a4c-aa00-6f5773d042c9","order_by":8,"name":"Zhihui Lu","email":"","orcid":"","institution":"Research Institute of Intelligent Control and Systems, Harbin Institute of Technology, Harbin, China","correspondingAuthor":false,"prefix":"","firstName":"Zhihui","middleName":"","lastName":"Lu","suffix":""},{"id":588395992,"identity":"6002da1c-f46a-4942-8a54-62479707ed45","order_by":9,"name":"Mingsi Tong","email":"","orcid":"https://orcid.org/0000-0002-1939-0040","institution":"Harbin Institute of Technology","correspondingAuthor":false,"prefix":"","firstName":"Mingsi","middleName":"","lastName":"Tong","suffix":""},{"id":588395993,"identity":"f1cb653b-7916-4a28-88a7-1d06563903e7","order_by":10,"name":"Li Zhang","email":"","orcid":"https://orcid.org/0000-0003-1152-8962","institution":"The Chinese University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2026-01-16 10:45:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8617009/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8617009/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102397614,"identity":"a02420b6-1833-40e1-8af0-af5b4f8ba478","added_by":"auto","created_at":"2026-02-11 10:18:27","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":10342078,"visible":true,"origin":"","legend":"Article File","description":"","filename":"nmmanuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8617009/v1_covered_e48a6cea-11b2-4c8a-b704-be41258bfedd.pdf"},{"id":102307732,"identity":"4c9ac8d3-ab44-455b-9f08-2498c24e6bfe","added_by":"auto","created_at":"2026-02-10 11:46:06","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1281341,"visible":true,"origin":"","legend":"Supplementary information","description":"","filename":"nmsupplementaryinformation.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8617009/v1/129509dbb00ef1d9216797c4.pdf"},{"id":102305358,"identity":"636b7725-8a0d-4ec4-a6a7-51ba80a2a03f","added_by":"auto","created_at":"2026-02-10 11:38:33","extension":"mp4","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":91825752,"visible":true,"origin":"","legend":"Overall introduction and multi-step experiment with fluorescence slice data acquisition and analysis.","description":"","filename":"SupplementaryVideo1.mp4","url":"https://assets-eu.researchsquare.com/files/rs-8617009/v1/05f16dbe93a28b2c77999bb8.mp4"},{"id":102305294,"identity":"6f2be7d5-e376-41d4-89a2-a2c040a6d578","added_by":"auto","created_at":"2026-02-10 11:38:16","extension":"mp4","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":130321347,"visible":true,"origin":"","legend":"Error correction experiment with fluorescence slice data acquisition and analysis.","description":"","filename":"SupplementaryVideo2.mp4","url":"https://assets-eu.researchsquare.com/files/rs-8617009/v1/e76ac7908f6762d64e7672dd.mp4"},{"id":102307850,"identity":"a990cda3-d3a5-4aea-828e-9d1a37ddc1b1","added_by":"auto","created_at":"2026-02-10 11:46:26","extension":"mp4","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":10136609,"visible":true,"origin":"","legend":"Cross-platform generalization experiments.","description":"","filename":"SupplementaryVideo3.mp4","url":"https://assets-eu.researchsquare.com/files/rs-8617009/v1/560579da4839f55def295ba2.mp4"},{"id":102305348,"identity":"0bade528-7328-4b87-ae11-bfb1e7c5ceac","added_by":"auto","created_at":"2026-02-10 11:38:29","extension":"mp4","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":5606482,"visible":true,"origin":"","legend":"Comparison with alternative embodied AI architectures.","description":"","filename":"SupplementaryVideo4.mp4","url":"https://assets-eu.researchsquare.com/files/rs-8617009/v1/d53d0c207080ad107d036cbc.mp4"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Embodied Intelligence Enables Agentic Exploration in Microscopy","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8617009/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8617009/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Microscopy has fundamentally transformed our understanding of life; however, despite relentless technological iteration, its underlying operational paradigm remains strikingly static. Current instruments persist as passive data collectors, rigidly executing predefined commands while remaining blind to the stochastic and dynamic nature of biological systems. This reliance on passive observation creates a semantic gap between high-level scientific inquiry and physical execution, limiting the generalizability in adapting to uncertain tasks or dynamic biological events. To bridge this gap, we present an Embodied Intelligence Microscope System (EIMS) that reconceptualizes the microscope as an autonomous robot capable of active exploration. By physically grounding the reasoning capabilities of intelligent system, EIMS closes the loop between scientific intent, physical perception, and action, enabling a transition from hypothesis-driven verification to discovery-driven explore. The system achieves zero-shot generalization across complex, multi-step tasks and can autonomously formulate experimental protocols, navigate unknown spatiotemporal environments to locate regions of interest, and dynamically optimizes imaging parameters in real-time to resolve ambiguity. This work establishes a new paradigm for self-driving instruments, providing a generalizable framework where embodied intelligence acts not merely as a tool, but as an active partner in autonomous scientific discovery.","manuscriptTitle":"Embodied Intelligence Enables Agentic Exploration in Microscopy","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-10 11:07:10","doi":"10.21203/rs.3.rs-8617009/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-methods","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nmeth","sideBox":"Learn more about [Nature Methods](http://www.nature.com/nmeth)","snPcode":"","submissionUrl":"","title":"Nature Methods","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"2b0bbb07-20ab-4606-a09c-4754242e446d","owner":[],"postedDate":"February 10th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":62598806,"name":"Biological sciences/Biological techniques/Software"},{"id":62598807,"name":"Biological sciences/Biological techniques/Microscopy/Confocal microscopy"}],"tags":[],"updatedAt":"2026-03-09T21:15:25+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-10 11:07:10","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8617009","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8617009","identity":"rs-8617009","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}