Utilization of Digital Twin (DT) and Artificial Intelligence (AI) for Building Decarbonization from a Life Cycle Perspective
Deniz Besiktepe (PI of the REU Site)

Background: The built environment accounts for approximately 45% of global carbon emissions and comprises almost 75% of building operating costs. Cutting-edge technologies such as DT and AI are promising for synthesizing big data from sensors, automation systems, maintenance activities, and building occupancy to transform building operations phase into a data-driven environment. Considering these, this study aims to develop a DT and AI based model (Figure 4) to predict the impact of failures and defects of critical building systems on carbon emissions, which provides a circular environment to improve decision-making for the operational efficiency of buildings. The research is driven by the PI Besiktepe’s current project in NSF I-Corps program (Award #2428909).
Learning Objectives: Students engaged in this project will be able to (a) utilize a digital twin model of a building with exploring the data and information structure, (b) analyze the sensor, automation, energy, and metering data of the building via use case, (c) explore the potential of artificial intelligence in data analytics and predictions, and (d) examine the impact of built environments operational phase on the life cycle and carbon footprint. This project will also cultivate students’ capability in creative thinking, research problem identification and definition, research methodology design, design of experimentation, results interpretation, and communication.
Research Approach: Students will explore the life-cycle carbon footprint using a DT and AI model. Specifically, explore hands-on DT modeling and AI techniques, and analyze the data from various resources in the buildings operational phase. Finally, they will examine the impact of building operations in the overall lifecycle from the decarbonization perspective, which will provide them to improve their awareness on the need for circularity approach on AECO industry rather than traditional linear models.
Decarbonization in Urban Built Environment from a System-Level Perspective
Soowon Chang (Co-PI of the REU Site)

Background: Energy infrastructure transitions involving rooftop photovoltaics (PV) and electric vehicles (EVs) have emerged as promising solutions to effectively decarbonize urban energy systems (Chang et al., 2022). However, there remains uncertainty regarding the collective contributions of these technologies to decarbonizing urban energy systems, particularly considering varying spatial conditions such as built forms, building types, or urban density (Figure 5). This project intends to support decisions of solar PV sizing in building assets, and EV adoption in transportation for sustainable and decarbonized civil systems by preparing the next generation of engineering equipped with life-cycle perspectives. The research is driven by the co PI Chang’s current project in NSF Build and Broaden program (Award # 2315876).
Learning Objectives: Students will (a) develop an appreciation for the interdependencies between the building and transportation sectors in the pursuit of decarbonizing communities, (b) collect socioeconomic data and potential low-carbon electricity potential data from diverse built environments, (c) recognize different geospatial scale perspectives, and (d) apply techno-economic analysis to assess the penetration rates of PV combined with EV energy systems.
Research Approach: By providing undergraduate researchers with system-level perspectives and advanced technical skills, this project aims to nurture the next generation of construction engineers and managers capable of addressing future-proof applications of PV on buildings and EV adoption rate in transportation. Undergraduate researchers will explore the decarbonization potential of rooftop PV and EV in various urban settings utilizing techno-economic analysis and a life-cycle perspective. The participating students will collect data on sociotechnical systems for local communities in the State of Indiana using a publicly accessible dataset (Indiana Geographic Information Officer (GIO), n.d.). They will learn geospatial data analytics using Geographic Information System (GIS) tools (e.g., ArcGIS and QGIS) and techno-economic analysis using System Advisory Model (SAM) (Freeman et al., 2018). Finally, they will implement these techniques to envision low environmental impact development for future sustainable communities. REU students will contribute to the development of process-based cost and system-level assessment models of low-carbon electricity applications.
Portable CO2 Emission Monitoring Device for Civil Infrastructure

Background: CO2 emission is the major contributor to global warming. Among all the sources, the construction industry generates more than one-third of the gross CO2 emission in the US. To achieve the carbon-neutral goal by 2050, it is critical to develop a strategic plan for reducing CO2 emissions. With a growing demand for constructing civil infrastructure, it is increasingly important to understand the life-cycle CO2 conditions of the infrastructure and monitor the CO2 conditions in real time.
Learning Objectives: Students will be able to (a) become familiar with CO2 emission monitoring in civil infrastructure (b) develop and assemble simple Internet-of-Things device, (c) conduct physical measurements of the CO2 level, and (d) analyze the CO2 density data and understand the CO2 distribution over time .
Research Approach: In the project, the student will assist the faculty in developing a novel portable device monitoring CO2 density for different conditions. The student will work with the faculty to install the device, collect CO2 density, and validate the device's performance under different scenarios. The collected data will be used to analyze the CO2 conditions and provide suggestions for improving civil infrastructure decarbonization. This project will help better monitor the CO2 conditions, investigate the CO2 emission of civil infrastructure in real time, and develop a better strategy for decarbonization. This project is part an effort funded by The Transportation Infrastructure Precast Innovation Center (TRANS-IPIC) a UTC federally funded by the USDOT.
Decarbonization of Cement and Concrete Materials for Better Sustainability

Background: Construction material industry is one of the largest and most energy and carbon intensive industries in the world. The cement and concrete industry are responsible for 1,500 megatons/year of CO2 emissions (approximately 0.8 - 1.0 ton of CO2 per ton of cement). It is significant to evaluate the environmental impact and decrease the carbon footprint for cement and concrete materials through its life cycle encompassing extracting and processing of the raw materials, manufacturing, distribution, installation, recycling, and final disposal. The research is supported by NSF I-Corps program and the American Concrete Institute (ACI) - Concrete Research Council (CRC).
Learning Objectives: Students will be able to (a) evaluate the environmental impact from cement and concrete material industry, (b) understand the life cycle process of construction materials from manufacturing, distribution, installation, recycling and final disposal, (c) model the cement and concrete materials manufacturing and hydration process, and (d) develop computational tools for cement and concrete decarbonization and life-cycle assessment.
Research Approach: The project will contribute to build a low-carbon, low-cost, and high-performance construction material industry. In the project, students will support the materials modeling, experimental validation, life cycle analysis for cement and concrete decarbonization. The student will work with the faculty to simulate the cement and concrete hydration process through Virtual Cement and Concrete Testing Laboratory (VCCTL) software, collect experimental data of the cement and concrete compressive strength in the soils and concrete lab at the School of Construction Management Technology, and evaluate environmental impact of different types of cement and concrete materials with LCA software.
Life Cycle Assessment of Construction Materials for Decarbonization using Athena Impact Estimator

Background: The construction industry significantly contributes to global carbon emissions, with a large portion arising from the embodied carbon of materials used throughout a building’s life cycle. Understanding the environmental impact of material selection and design decisions early in the process is crucial for achieving sustainable and low-carbon construction. This project introduces undergraduate researchers to Life Cycle Assessment (LCA) as a method for quantifying environmental impacts associated with building materials and assemblies. Using the Athena Impact Estimator for Buildings, students will model and compare the embodied carbon of different materials to evaluate how early design choices influence a building’s carbon footprint and explore practical strategies for reducing emissions.
Learning Objectives: By the end of this research experience, students will have developed a comprehensive understanding of Life Cycle Assessment (LCA) principles and their relevance to sustainable construction and decarbonization goals. They will gain practical skills in using the Athena Impact Estimator for Buildings to model and analyze the environmental impacts of various construction materials and assemblies. Through the interpretation and comparison of LCA results, students will learn to identify materials and design strategies that contribute most significantly to embodied carbon reduction. Additionally, they will enhance their ability to communicate research findings effectively through technical posters, written reports, and symposium presentations. This experience will also strengthen their critical thinking and problem-solving abilities within the context of sustainability-oriented design and decision-making in the construction industry.
Research Approach: The research will be conducted through a structured process that integrates training, modeling, analysis, and dissemination. Students will begin by gaining a foundational understanding of sustainability concepts and Life Cycle Assessment (LCA) methodology, along with hands-on training in the use of the Athena Impact Estimator for Buildings. Building on this foundation, they will develop and refine LCA models for various construction materials and assemblies, focusing on quantifying embodied carbon and energy use. The analysis phase will involve comparing results to identify materials, assemblies, and design decisions that have the greatest influence on a building’s overall environmental impact. Finally, students will synthesize their findings into a research poster and written report that connect their results to broader decarbonization pathways within the built environment, culminating in a formal presentation at the Purdue OUR Summer Symposium.
Sociotechnical Data Framework for Evaluating Energy Interventions

Background: Smart energy assistants can be a powerful tool for households to monitor their energy use, use energy efficiently, and maximize the benefits of their home’s heating and cooling equipment. Despite these benefits, low-income households have been slower to adopt these technologies. This NSF-funded project (Award #2331940) aims to collaborate with residents of affordable housing communities and housing agencies to learn more about the best ways to engage these types of residents via smart energy devices and use machine learning techniques to provide recommendations that balance resident comfort with energy savings, especially when using electrified heating and cooling systems.
Learning Objectives: Students engaged in this project will become familiar with energy use data collected with smart energy assistants, weather data, and survey data related to energy assistants. Students will be able to use this data to categorize residents by their energy use patterns and identify strategies for engaging them to use energy assistants to reduce their energy use.
Research Approach: Students will synthesize data from multiple sources (weather databases, surveys, smart energy assistants) and use this information to estimate the energy consumption for a typical home under the specified weather conditions. Counterfactual scenarios will incorporate building information, electricity price information, and engagement strategies to model potential impacts of energy assistant communication strategies.
Electrical Vehicle Maintenance and Repair through Immersive Reality

Background: With the rapid adoption of electric vehicles (EVs), there will be numerous new job opportunities for EV maintenance and repairs for the next several decades. EV maintenance and repair require professional knowledge from multiple domains, which makes it challenging for existing training methods to create immersive and effective learning experiences for automotive technicians.
Learning Objectives: Students will be able to (a) evaluate the use case of EV with its benefits on decarbonization, (b) understand the principles of immersive reality with utilizing Trimble equipment, (c) determine the importance of the life cycle approach on EV with emphasizing the systems approach, and (d) experience the development of a maintenance and repair plan for EVs.
Research Approach: The project will contribute to making connections between EVs and their benefits on addressing the global greenhouse gas emissions problem. In the project, students will support the faculty on implementing the immersive reality tools with simulations on EVs. The investigation for the maintenance and repair plan for an EV would provide a better understanding on the holistic and systems level approach on the decarbonization in built environment. The student will also work with the project’s team in a collaborative environment, gaining experience from different universities and benefiting from the diverse perspectives and expertise of team members from various academic backgrounds. This project is part of an effort funded by Dr. Hasanzadeh’s current NSF Award # 2347196.
AI-Driven Vision Inspection Analytics to Assist with Quality Control of Bridge Inspection Documentation

Background: Bridges are critical components of transportation infrastructure, yet their inspections often rely on subjective visual assessments that can lead to inconsistent documentation and delayed maintenance decisions. To address this challenge, the project AI-Driven Vision Inspection Analytics to Assist with Quality Control of Bridge Inspection Documentation, funded by the Indiana Department of Transportation (INDOT), develops an artificial intelligence system that leverages over 42,000 bridge inspection images and reports. Using deep learning and vision-language models, the system detects defects, verifies inspection quality, and provides automated feedback to improve inspection consistency and asset longevity. By enabling data-driven maintenance and extending bridge service life, this project contributes to built environment decarbonization through reduced material waste, minimized reconstruction needs, and more sustainable infrastructure management.
Learning Objectives: Students participating in this project will gain hands-on experience at the intersection of artificial intelligence, computer vision, and civil infrastructure management. They will learn how to preprocess and analyze large-scale image and text datasets derived from real bridge inspection reports, apply deep learning and vision-language models for defect detection and documentation quality control, and evaluate AI model performance in engineering contexts. In addition, students will develop skills in data annotation, model training, and prototype testing using field-relevant datasets. Through this work, they will understand how AI can support sustainable infrastructure practices and contribute to the broader goals of built environment decarbonization by promoting data-driven maintenance and extending the lifespan of critical assets.
Research Approach: The research approach integrates artificial intelligence and computer vision to enhance bridge inspection quality and sustainability. Students will assist in compiling and preprocessing INDOT’s archive of over 42,000 bridge inspection images and reports to create labeled datasets for training AI models. Using deep learning and vision-language architectures, the team will develop and test hybrid models capable of detecting structural defects, verifying inspection documentation, and providing interpretable feedback. The validated model will be incorporated into a user-friendly interface for inspectors, enabling real-time quality control. This data-driven approach supports efficient maintenance decisions and contributes to the decarbonization of the built environment by reducing unnecessary reconstruction and material use.