Search and Explore: Symbiotic Policy Synthesis in POMDPs

Search and Explore: Symbiotic Policy Synthesis in POMDPs | 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 Research Article Search and Explore: Symbiotic Policy Synthesis in POMDPs Roman Andriushchenko, Alexander Bork, Milan Češka, Sebastian Junges, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4363964/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 20 Feb, 2026 Read the published version in Formal Methods in System Design → Version 1 posted 9 You are reading this latest preprint version Abstract This paper marries two state-of-the-art controller synthesis methods for partially observable Markov decision processes (POMDPs), a prominent model in sequential decision making under uncertainty. A central issue is to find a POMDP controller - that solely decides based on the observations seen so far - to achieve a total expected reward objective. As finding optimal controllers is undecidable, we concentrate on synthesising good finite-state controllers (FSCs). We do so by tightly integrating two modern, orthogonal methods for POMDP controller synthesis: a belief-based and an inductive approach. The former method obtains an FSC from a finite fragment of the so-called belief MDP, an MDP that keeps track of the probabilities of equally observable POMDP states. The latter is an inductive search technique over a set of FSCs, e.g., controllers with a fixed memory size. The key result of this paper is a symbiotic anytime algorithm that tightly integrates both approaches such that each profits from the controllers constructed by the other. Experimental results indicate a substantial improvement in the value of the controllers while significantly reducing the synthesis time and memory footprint. Partially observable MDPs belief space finite-state controllers inductive synthesis Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 20 Feb, 2026 Read the published version in Formal Methods in System Design → Version 1 posted Editorial decision: Revision requested 23 Apr, 2025 Reviews received at journal 11 Apr, 2025 Reviewers agreed at journal 21 Mar, 2025 Reviews received at journal 08 Jul, 2024 Reviewers agreed at journal 13 May, 2024 Reviewers invited by journal 13 May, 2024 Editor assigned by journal 08 May, 2024 Submission checks completed at journal 07 May, 2024 First submitted to journal 03 May, 2024 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-4363964","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":301819098,"identity":"9ccbd47b-1f92-4416-b7b5-ec295543e87d","order_by":0,"name":"Roman Andriushchenko","email":"","orcid":"","institution":"Brno University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Roman","middleName":"","lastName":"Andriushchenko","suffix":""},{"id":301819100,"identity":"42fc440c-2195-414f-af7a-b6212f882d17","order_by":1,"name":"Alexander Bork","email":"","orcid":"","institution":"RWTH Aachen University","correspondingAuthor":false,"prefix":"","firstName":"Alexander","middleName":"","lastName":"Bork","suffix":""},{"id":301819101,"identity":"6d3733f4-6a76-4be9-b794-bc5f6fffceaa","order_by":2,"name":"Milan Češka","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwUlEQVRIiWNgGAWjYHACZiC24WcjVUuaZBtYTwLxWg5LNhCthb+9/bHBz5zzEnzyDWzSvD8Y5OUbCGiROHPGOLF3220JNjYGNmmeBAbDRkJaDCRymA/wbrtdB9TCbAzUksBMyGEG8s8fH/y77ZwEXAvBoDOQYDBO5t12AKSF8TFICw8hLRJncoyNZbclA7UkNj6ckyZhOIOQFv72448l326zk5BvPnzgwBsbG8IhhgQYQWoliFc/CkbBKBgFowA3AACTyDDuEX68kwAAAABJRU5ErkJggg==","orcid":"","institution":"Brno University of Technology","correspondingAuthor":true,"prefix":"","firstName":"Milan","middleName":"","lastName":"Češka","suffix":""},{"id":301819103,"identity":"0d37e7c4-47a5-49d6-b777-5259eb7cbc11","order_by":3,"name":"Sebastian Junges","email":"","orcid":"","institution":"Radboud University Nijmegen","correspondingAuthor":false,"prefix":"","firstName":"Sebastian","middleName":"","lastName":"Junges","suffix":""},{"id":301819107,"identity":"83d28d2a-a655-4984-895e-b5d61d5658ff","order_by":4,"name":"Joost-Pieter Katoen","email":"","orcid":"","institution":"RWTH Aachen University","correspondingAuthor":false,"prefix":"","firstName":"Joost-Pieter","middleName":"","lastName":"Katoen","suffix":""},{"id":301819110,"identity":"d4d2aa44-33be-49ec-a310-133de4308ef3","order_by":5,"name":"Filip Macák","email":"","orcid":"","institution":"Brno University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Filip","middleName":"","lastName":"Macák","suffix":""}],"badges":[],"createdAt":"2024-05-03 11:57:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4363964/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4363964/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s10703-025-00488-z","type":"published","date":"2026-02-20T15:58:41+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":103252088,"identity":"64a0398c-1faf-40cc-a840-9bf895757840","added_by":"auto","created_at":"2026-02-23 16:12:57","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":940375,"visible":true,"origin":"","legend":"","description":"","filename":"CAV23SISearchandExploreSymbioticPolicySynthesisinPOMDPs.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4363964/v1_covered_261e1b9f-c7ca-4f38-b33a-eb8354e0075a.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Search and Explore: Symbiotic Policy Synthesis in POMDPs","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"formal-methods-in-system-design","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"form","sideBox":"Learn more about [Formal Methods in System Design](https://www.springer.com/journal/10703)","snPcode":"10703","submissionUrl":"https://submission.springernature.com/new-submission/10703/3","title":"Formal Methods in System Design","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Partially observable MDPs, belief space, finite-state controllers, inductive synthesis","lastPublishedDoi":"10.21203/rs.3.rs-4363964/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4363964/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"This paper marries two state-of-the-art controller synthesis methods for partially observable Markov decision processes (POMDPs), a prominent model in sequential decision making under uncertainty. A central issue is to find a POMDP controller - that solely decides based on the observations seen so far - to achieve a total expected reward objective. As finding optimal controllers is undecidable, we concentrate on synthesising good finite-state controllers (FSCs). We do so by tightly integrating two modern, orthogonal methods for POMDP controller synthesis: a belief-based and an inductive approach. The former method obtains an FSC from a finite fragment of the so-called belief MDP, an MDP that keeps track of the probabilities of equally observable POMDP states. The latter is an inductive search technique over a set of FSCs, e.g., controllers with a fixed memory size. The key result of this paper is a symbiotic anytime algorithm that tightly integrates both approaches such that each profits from the controllers constructed by the other. Experimental results indicate a substantial improvement in the value of the controllers while significantly reducing the synthesis time and memory footprint.","manuscriptTitle":"Search and Explore: Symbiotic Policy Synthesis in POMDPs","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-05-14 17:42:22","doi":"10.21203/rs.3.rs-4363964/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-04-23T13:35:52+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-04-11T21:02:22+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"233416889328552583942447150477068285187","date":"2025-03-21T10:34:13+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-07-08T06:59:27+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"289064464864124603152351592077911309637","date":"2024-05-13T09:48:39+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-05-13T09:46:29+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-05-08T06:23:22+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-05-07T10:33:40+00:00","index":"","fulltext":""},{"type":"submitted","content":"Formal Methods in System Design","date":"2024-05-03T11:55:52+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"formal-methods-in-system-design","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"form","sideBox":"Learn more about [Formal Methods in System Design](https://www.springer.com/journal/10703)","snPcode":"10703","submissionUrl":"https://submission.springernature.com/new-submission/10703/3","title":"Formal Methods in System Design","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"9849a8e0-a789-42be-a46b-766059dc4587","owner":[],"postedDate":"May 14th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-02-23T16:08:47+00:00","versionOfRecord":{"articleIdentity":"rs-4363964","link":"https://doi.org/10.1007/s10703-025-00488-z","journal":{"identity":"formal-methods-in-system-design","isVorOnly":false,"title":"Formal Methods in System Design"},"publishedOn":"2026-02-20 15:58:41","publishedOnDateReadable":"February 20th, 2026"},"versionCreatedAt":"2024-05-14 17:42:22","video":"","vorDoi":"10.1007/s10703-025-00488-z","vorDoiUrl":"https://doi.org/10.1007/s10703-025-00488-z","workflowStages":[]},"version":"v1","identity":"rs-4363964","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4363964","identity":"rs-4363964","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}