[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-124248-en":3,"doc-seo-124248-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},124248,687197207057,"Sage","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","Automatic Pavement Maintenance Predictions using Large Language and Machine Learning Models","Pavement quality throughout its lifespan depends on effective maintenance and rehabilitation (M&R) planning, and evaluating past M&R records is critical for reliable future decisions. Timestamped sources are often incomplete in public repositories such as the Long-Term Pavement Performance (LTPP) database, especially when numerical and textual gaps coexist. This study develops a framework that uses ChatGPT-4 to classify major or minor M&R activities, then combines these labels with pavement numerical features for sequential time-series prediction. Machine learning models are trained to forecast major, minor, or no maintenance, with XGBoost achieving strong F1 performance on two- and three-year datasets.","Automatic Pavement Maintenance Predictions using Large Language and Machine  \nLearning Models.  \nKunle Sunday Oguntoye 1, Halil Ceylan 2, Sunghwan Kim 3, Md Abdullah All Sourav 4, Berk  \nGulmezoglu 5, Yunjeong Mo 6  \n1 Ph.D. Candidate, Department of Civil, Construction, and Environmental Engineering, Iowa State University, Ames, IA. Email: [oguntoye@iastate.edu](oguntoye@iastate.edu)  \n2 Professor, Department of Civil, Construction, and Environmental Engineering, and Director, Program for Sustainable Pavement Engineering and Research, Iowa State University, Ames, IA. [Email: ](Email: hceylan@iastate.edu)[hceylan@iastate.edu](Email: hceylan@iastate.edu)  \n3 Associate Director, Program for Sustainable Pavement Engineering and Research, Institute for Transportation, Iowa State University, Ames, IA. Email: [sunghwan@iastate.edu](sunghwan@iastate.edu)  \n4 Research Associate, Department of Civil, Construction, and Environmental Engineering, Iowa State University, Ames, IA. Email: [sourav@iastate.edu](sourav@iastate.edu)  \n5 Assistant Professor, Department of Electrical and Computer Engineering, Iowa State University, Ames, [IA. Email: ](IA. Email: bgulmez@iastate.edu)[bgulmez@iastate.edu](IA. Email: bgulmez@iastate.edu)  \n6 Assistant Professor, Department of Civil, Construction, and Environmental Engineering, Iowa State University, Ames, [IA. Email: ](IA. Email: ymo@iastate.edu)[ymo@iastate.edu](IA. Email: ymo@iastate.edu)  \nABSTRACT  \nPavement quality throughout its lifespan depends on effective maintenance and rehabilitation (M&R) . Evaluating past M&R efforts is essential for planning future activities. While wellorganized, timestamped records support preventive maintenance, public databases like the LongTerm Pavement Performance (LTPP) database often have incomplete M&R records. Missing numerical data, such as rutting, can be statistically imputed, but addressing gaps in textual M&R records is more complex. This study proposes a framework utilizing OpenAI’s ChatGPT-4 to classify M&R records as major or minor activities. The annotated records, combined with numerical pavement features like layer information and rutting, are structured into sequential time series data. Machine learning models—CatBoost, XGBoost, LightGBM, and decision tree classifiers—are trained to predict whether major, minor or no maintenance occurred. The XGBoost model outperformed others, achieving F1 scores of 82% and 89% for two-year and three-year time series datasets, offering a robust solution to incomplete pavement management data.  \nKeywords: maintenance and rehabilitation, large language models, machine learning models, pavement management systems, ChatGPT, time series, LTPP datasets.  \nINTRODUCTION  \nPavement maintenance and rehabilitation (M&R) activities are some of the sustainable programs targeted toward achieving long-lasting pavement infrastructure that maintains a high level of service. Early works by Carnahan (1988) identified the high cost of premature pavement replacement or rehabilitation resulting from inadequate routine maintenance activities, and it has since become a regular program to carry out both corrective and preventive maintenance for the preservation of infrastructure resiliency. However, drawbacks such as limited funding for  \nmaintenance programs have challenged researchers to develop optimized and continuous maintenance activities using prioritization models and outcomes from analyzing the benefit-cost of different maintenance activities in a pavement deterioration model (France-Mensah and O’Brien 2018) . For example, Ahmed et al. (2017) and Farhna and Fwa (2009) utilized an analytic hierarchy process that closely simulates the engineering judgment of highway agencies and engineers in ranking pavements in order of maintenance priority. Bandara and Gunaratne (2001) utilized a fuzzy aggregation system in ranking pavement maintenance priority using rapid expert visual condition evaluations. Alsugair and Al-Qudrah (1998) develope","cbCaipQV82S6hPMM","https://ap.wps.com/l/cbCaipQV82S6hPMM","pdf",600112,1,12,"English","en",105,"# Abstract\n# Introduction\n## Maintenance planning and sustainability\n## Predictive approaches and gaps in LTPP records\n## Benefits of using historical maintenance information","[{\"question\":\"Why are past maintenance and rehabilitation (M\\u0026R) records important for pavement management?\",\"answer\":\"Past M\\u0026R information supports planning future activities by guiding maintenance selection and helping engineers and managers track budgets across the pavement lifespan.\"},{\"question\":\"What problem does the study address in public pavement datasets like LTPP?\",\"answer\":\"The LTPP repository often contains incomplete M\\u0026R records, particularly missing textual information, making it difficult to model maintenance history accurately.\"},{\"question\":\"How does the proposed framework handle missing textual M\\u0026R records?\",\"answer\":\"The framework uses ChatGPT-4 to classify records into major or minor activities, then merges these classifications with numerical pavement features to train time-series prediction models.\"}]","Automatic Pavement Maintenance Predictions using Large Language and Machine Learning Models | PDF",1785821220,30,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"automatic-pavement-maintenance-predictions-using-large-language-and-machine-learning-models","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/automatic-pavement-maintenance-predictions-using-large-language-and-machine-learning-models/124248/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-04",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why are past maintenance and rehabilitation (M&R) records important for pavement management?","Question",{"text":75,"@type":76},"Past M&R information supports planning future activities by guiding maintenance selection and helping engineers and managers track budgets across the pavement lifespan.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What problem does the study address in public pavement datasets like LTPP?",{"text":80,"@type":76},"The LTPP repository often contains incomplete M&R records, particularly missing textual information, making it difficult to model maintenance history accurately.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the proposed framework handle missing textual M&R records?",{"text":84,"@type":76},"The framework uses ChatGPT-4 to classify records into major or minor activities, then merges these classifications with numerical pavement features to train time-series prediction models.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":29,"slug":121},"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]