HyFiD: LLM-ML Hybrid Framework for Subway Fire Detection

HyFiD: LLM-ML Hybrid Framework for Subway Fire Detection | 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 HyFiD: LLM-ML Hybrid Framework for Subway Fire Detection Kihwan Ko, Ikgeun Kwon, Yujin Kang, Hee-Dong Kim, Yoon-Sik Cho This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8760486/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 12 You are reading this latest preprint version Abstract Fire detection is vital in subway tunnels where confined geometries and ventilation complicate safety monitoring. Existing methods, such as classical machine learning, detect fire based on multivariate correlations but often lack the contextual reasoning required for disambiguation. This limitation becomes critical when HVAC-driven airflow disrupts thermal stratification and dilutes gas concentrations, creating ambiguous patterns that mimic fire signatures. Recent studies further suggest that Large Language Models (LLMs) can overcome this bottleneck by leveraging pretrained knowledge to interpret complex sensor dynamics into concise, semantic descriptions. To effectively integrate this semantic capability into a robust detection framework, we propose HyFiD, a hybrid framework that employs an LLM as a semantic feature extractor to augment classical classifiers. By converting momentary multi-sensor readings (temperature, smoke, O2, CO, and CO2) into textual assessments of environmental states, HyFiD generates semantic vectors that are fused with numerical features for robust training. Experiments on Fire Dynamics Simulator (FDS)-based subway scenarios, including ventilation-dominated HVAC events and high-energy battery fires, reveal that semantic augmentation improves performance over numerical-only baselines and prompt-only LLM classifiers, while enhancing interpretability through explicit, human-readable evidence. Physical sciences/Energy science and technology Physical sciences/Engineering Physical sciences/Mathematics and computing Fire Detection Subway Fire Safety Large Language Models Machine Learning Prompt Engineering Hybrid Framework Fire Dynamics Simulator Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 01 Apr, 2026 Reviews received at journal 24 Mar, 2026 Reviews received at journal 17 Mar, 2026 Reviewers agreed at journal 28 Feb, 2026 Reviewers agreed at journal 26 Feb, 2026 Reviewers agreed at journal 26 Feb, 2026 Reviewers agreed at journal 26 Feb, 2026 Reviewers invited by journal 26 Feb, 2026 Editor invited by journal 24 Feb, 2026 Editor assigned by journal 03 Feb, 2026 Submission checks completed at journal 03 Feb, 2026 First submitted to journal 01 Feb, 2026 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-8760486","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":598586288,"identity":"94cee1f0-488e-4187-97fa-d21a15862004","order_by":0,"name":"Kihwan Ko","email":"","orcid":"","institution":"Chung-Ang University","correspondingAuthor":false,"prefix":"","firstName":"Kihwan","middleName":"","lastName":"Ko","suffix":""},{"id":598586289,"identity":"08fed67b-e20e-42a3-86f9-4f32549a88f2","order_by":1,"name":"Ikgeun Kwon","email":"","orcid":"","institution":"Sejong University","correspondingAuthor":false,"prefix":"","firstName":"Ikgeun","middleName":"","lastName":"Kwon","suffix":""},{"id":598586290,"identity":"3eaec985-2c21-4502-acc9-ce84c0539b00","order_by":2,"name":"Yujin Kang","email":"","orcid":"","institution":"Chung-Ang University","correspondingAuthor":false,"prefix":"","firstName":"Yujin","middleName":"","lastName":"Kang","suffix":""},{"id":598586291,"identity":"08ff0f0c-b57a-4776-969b-cbe0bb1b879e","order_by":3,"name":"Hee-Dong Kim","email":"","orcid":"","institution":"Sejong University","correspondingAuthor":false,"prefix":"","firstName":"Hee-Dong","middleName":"","lastName":"Kim","suffix":""},{"id":598586292,"identity":"816745b2-5359-4e1f-8382-5bc991bd7173","order_by":4,"name":"Yoon-Sik Cho","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwklEQVRIiWNgGAWjYNACgwQ5OJNoLcakamFISGwgWovB8eZjD38UpKX3t58xYPhRw2Bs3kBIy5lj6cY8Bjm5M87kGDD2HGMwkzlASMuNHDNpBoOK3A0MOQYMvA0MNhIEHQbUIvnDoCLdgP+NAeNfYrVIAB2WYCCRY8AMtMWMoBbJM8fSpHkM0gxn3HhWcFjmmIQxQS18wBCT/PEnWZ6/P3njwzc1NoYzCGlROIDEAbIJ2sHAIN9AWM0oGAWjYBSMdAAAhF45OtBFK7MAAAAASUVORK5CYII=","orcid":"","institution":"Chung-Ang University","correspondingAuthor":true,"prefix":"","firstName":"Yoon-Sik","middleName":"","lastName":"Cho","suffix":""}],"badges":[],"createdAt":"2026-02-02 04:53:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8760486/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8760486/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":104400935,"identity":"93b7603f-2fa7-4fbb-b940-fda0396a880a","added_by":"auto","created_at":"2026-03-11 12:11:28","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":621057,"visible":true,"origin":"","legend":"","description":"","filename":"HyFiD0203.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8760486/v1_covered_c2d40804-1c4a-45ac-8826-fd4adef7269c.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eHyFiD: LLM-ML Hybrid Framework for Subway Fire Detection\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"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":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Fire Detection, Subway Fire Safety, Large Language Models, Machine Learning, Prompt Engineering, Hybrid Framework, Fire Dynamics Simulator","lastPublishedDoi":"10.21203/rs.3.rs-8760486/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8760486/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Fire detection is vital in subway tunnels where confined geometries and ventilation complicate safety monitoring. Existing methods, such as classical machine learning, detect fire based on multivariate correlations but often lack the contextual reasoning required for disambiguation. This limitation becomes critical when HVAC-driven airflow disrupts thermal stratification and dilutes gas concentrations, creating ambiguous patterns that mimic fire signatures. Recent studies further suggest that Large Language Models (LLMs) can overcome this bottleneck by leveraging pretrained knowledge to interpret complex sensor dynamics into concise, semantic descriptions. To effectively integrate this semantic capability into a robust detection framework, we propose HyFiD, a hybrid framework that employs an LLM as a semantic feature extractor to augment classical classifiers. By converting momentary multi-sensor readings (temperature, smoke, O2, CO, and CO2) into textual assessments of environmental states, HyFiD generates semantic vectors that are fused with numerical features for robust training. Experiments on Fire Dynamics Simulator (FDS)-based subway scenarios, including ventilation-dominated HVAC events and high-energy battery fires, reveal that semantic augmentation improves performance over numerical-only baselines and prompt-only LLM classifiers, while enhancing interpretability through explicit, human-readable evidence.","manuscriptTitle":"HyFiD: LLM-ML Hybrid Framework for Subway Fire Detection","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-03 11:31:16","doi":"10.21203/rs.3.rs-8760486/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-01T05:06:41+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-24T04:50:34+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-18T02:38:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"81159463543395407037188447035158667195","date":"2026-02-28T15:52:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"217534460561407462904559879004598952325","date":"2026-02-26T15:15:26+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"307439262552395025018018664963312794695","date":"2026-02-26T15:03:21+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"161669476392607333204975781959951907060","date":"2026-02-26T15:03:20+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-02-26T14:51:32+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-02-24T12:24:39+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-02-04T01:25:45+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-02-04T01:25:43+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-02-02T04:41:49+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"325bbd19-73f5-4223-9cbd-55ac1f0f5137","owner":[],"postedDate":"March 3rd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":63841764,"name":"Physical sciences/Energy science and technology"},{"id":63841765,"name":"Physical sciences/Engineering"},{"id":63841766,"name":"Physical sciences/Mathematics and computing"}],"tags":[],"updatedAt":"2026-05-02T07:40:09+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-03 11:31:16","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8760486","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8760486","identity":"rs-8760486","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}