{"id":{"repo_id":"msu","oai_identifier":"oai:d.lib.msu.edu:etd_53293"},"canonical_url":"https://search.ndltd.org/etd/msu/oai:d.lib.msu.edu:etd_53293","repository":{"repo_id":"msu","name":"Michigan State University","base_url":"https://d.lib.msu.edu/oai"},"display":{"title":"AI-ENABLED RESIDENTIAL ENERGY RETROFIT DECISION SUPPORT FOR HOMEOWNERS AND COMMUNITY STAKEHOLDERS","abstract":"Thesis (Ph.D.)--Michigan State University. Planning, Design and Construction - Doctor of Philosophy, 2026","abstract_html":"Thesis (Ph.D.)--Michigan State University. Planning, Design and Construction - Doctor of Philosophy, 2026","abstract_has_math":false,"creators":["Shu, Lei"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["Zhao, Dong","Mollaoglu, Sinem","Zhou, Jiayu","Yeganeh, Armin"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026","date_published":"2026","updated_at":"2026-11-18T18:59:44Z","subjects":["Civil engineering"],"languages":["eng"],"rights":["In Copyright"],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["etd:53293","local:Shu_grad.msu_0128D_22345"],"render_values":[{"text":"etd:53293","href":null,"code":true},{"text":"local:Shu_grad.msu_0128D_22345","href":null,"code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":"https://doi.org/doi:10.25335/6j53-5c48","outbound_label":"DOI","outbound_source":"dc:identifier","record_url":"https://doi.org/doi:10.25335/6j53-5c48","record_label":"Repository record","record_kind":"doi","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://d.lib.msu.edu/oai?verb=GetRecord&metadataPrefix=oai_dc&identifier=oai%3Ad.lib.msu.edu%3Aetd_53293","prefix":"oai_dc"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Zhao, Dong","Mollaoglu, Sinem","Zhou, Jiayu","Yeganeh, Armin"]},{"key":"dc:creator","label":"Author","values":["Shu, Lei"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2026"]},{"key":"dc:relation","label":"Dc Relation","values":["Electronic Theses & Dissertations"]},{"key":"dc:type","label":"Dc Type","values":["text","Electronic dissertations"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Civil engineering"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["In Copyright"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["etd:53293","local:Shu_grad.msu_0128D_22345","https://doi.org/doi:10.25335/6j53-5c48"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Thesis (Ph.D.)--Michigan State University. Planning, Design and Construction - Doctor of Philosophy, 2026","Residential energy retrofits play a critical role in reducing building energy consumption, greenhouse gas emissions, and household energy costs. However, residential retrofit decision-making remains challenging because traditional assessment approaches often require detailed building information, specialized expertise, substantial time and computational resources. These requirements create accessibility barriers for non-expert stakeholders and limit the scalability of retrofit assessment across large building stocks.To address these challenges, this dissertation develops a series of artificial intelligence (AI)-enabled approaches for residential retrofit decision support at both the building and community levels. The research integrates physics-based simulation data, machine learning (ML), large language models (LLMs), techno-economic evaluation, and policy information to support informed retrofit decision-making while reducing input information requirements. The first study develops a domain-specific LLM for homeowner-oriented retrofit decision support. Using a dataset derived from physics-based simulations and techno-economic calculations for over half a million U.S. residential building prototypes, the model was fine-tuned to provide retrofit recommendations from homeowner-accessible natural-language descriptions. The model successfully identified high-quality retrofit recommendations, with the optimal retrofit measure included among the three recommended alternatives in 98.9% of cases for maximum CO2 reduction and 93.3% of cases for the shortest discounted payback year (DPY), while maintaining robust performance under incomplete input conditions. The second study develops an ML-based retrofit assessment framework that enables rapid retrofit evaluation using a limited set of user-provided building characteristics. Through feature–retrofit guidance and prototype-based input completion, the framework reduces information requirements while maintaining reliable predictive performance. The results showed 90% prediction accuracy for both optimal retrofit selection and DPY category prediction, while validation against real-world retrofit cases demonstrated a 90% agreement rate with expert-informed decisions. Building upon the building-level studies, the third study develops an integrated community-level retrofit decision-support framework. The framework combines ML-based retrofit performance prediction, policy-adjusted evaluation, and LLM-based recommendation generation. Even when only limited community information was available, the prediction models achieved R2 values of 0.97 for retrofit cost and 0.90 for CO2 reduction. The framework further incorporates policy incentives and automatically generates stakeholder-oriented recommendation reports to support community-scale retrofit planning. Collectively, the findings demonstrate that AI-enabled approaches can improve both the accessibility and scalability of residential retrofit decision-making. By reducing information barriers, supporting reliable retrofit assessment, and translating technical results into accessible recommendations, the proposed approaches provide practical pathways for homeowners, and community stakeholders to support residential building decarbonization.","Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).","Includes bibliographical references."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf","1 online resource (114 p.) : ill."]},{"key":"dc:title","label":"Title","values":["AI-ENABLED RESIDENTIAL ENERGY RETROFIT DECISION SUPPORT FOR HOMEOWNERS AND COMMUNITY STAKEHOLDERS"]}]}],"canonical_facts":{"dc:contributor":["Zhao, Dong","Mollaoglu, Sinem","Zhou, Jiayu","Yeganeh, Armin"],"dc:creator":["Shu, Lei"],"dc:date":["2026"],"dc:description":["Thesis (Ph.D.)--Michigan State University. Planning, Design and Construction - Doctor of Philosophy, 2026","Residential energy retrofits play a critical role in reducing building energy consumption, greenhouse gas emissions, and household energy costs. However, residential retrofit decision-making remains challenging because traditional assessment approaches often require detailed building information, specialized expertise, substantial time and computational resources. These requirements create accessibility barriers for non-expert stakeholders and limit the scalability of retrofit assessment across large building stocks.To address these challenges, this dissertation develops a series of artificial intelligence (AI)-enabled approaches for residential retrofit decision support at both the building and community levels. The research integrates physics-based simulation data, machine learning (ML), large language models (LLMs), techno-economic evaluation, and policy information to support informed retrofit decision-making while reducing input information requirements. The first study develops a domain-specific LLM for homeowner-oriented retrofit decision support. Using a dataset derived from physics-based simulations and techno-economic calculations for over half a million U.S. residential building prototypes, the model was fine-tuned to provide retrofit recommendations from homeowner-accessible natural-language descriptions. The model successfully identified high-quality retrofit recommendations, with the optimal retrofit measure included among the three recommended alternatives in 98.9% of cases for maximum CO2 reduction and 93.3% of cases for the shortest discounted payback year (DPY), while maintaining robust performance under incomplete input conditions. The second study develops an ML-based retrofit assessment framework that enables rapid retrofit evaluation using a limited set of user-provided building characteristics. Through feature–retrofit guidance and prototype-based input completion, the framework reduces information requirements while maintaining reliable predictive performance. The results showed 90% prediction accuracy for both optimal retrofit selection and DPY category prediction, while validation against real-world retrofit cases demonstrated a 90% agreement rate with expert-informed decisions. Building upon the building-level studies, the third study develops an integrated community-level retrofit decision-support framework. The framework combines ML-based retrofit performance prediction, policy-adjusted evaluation, and LLM-based recommendation generation. Even when only limited community information was available, the prediction models achieved R2 values of 0.97 for retrofit cost and 0.90 for CO2 reduction. The framework further incorporates policy incentives and automatically generates stakeholder-oriented recommendation reports to support community-scale retrofit planning. Collectively, the findings demonstrate that AI-enabled approaches can improve both the accessibility and scalability of residential retrofit decision-making. By reducing information barriers, supporting reliable retrofit assessment, and translating technical results into accessible recommendations, the proposed approaches provide practical pathways for homeowners, and community stakeholders to support residential building decarbonization.","Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).","Includes bibliographical references."],"dc:format":["application/pdf","1 online resource (114 p.) : ill."],"dc:identifier":["etd:53293","local:Shu_grad.msu_0128D_22345","https://doi.org/doi:10.25335/6j53-5c48"],"dc:language":["eng"],"dc:relation":["Electronic Theses & Dissertations"],"dc:rights":["In Copyright"],"dc:subject":["Civil engineering"],"dc:title":["AI-ENABLED RESIDENTIAL ENERGY RETROFIT DECISION SUPPORT FOR HOMEOWNERS AND COMMUNITY STAKEHOLDERS"],"dc:type":["text","Electronic dissertations"]},"updated_at":"2026-11-18T18:59:44Z"}