About

Amir Behzadan is a professor in the Department of Civil, Environmental and Architectural Engineering at the University of Colorado Boulder. He is also a fellow in the Institute of Behavioral Science (IBS), where the Natural Hazards Center is located. Behzadan is the director of the Connected Informatics and Built Environment Research (CIBER) Lab, which investigates grand challenges at the intersection of society and built/natural environments, including disaster resilience, climate change adaptation, jobsite safety, workplace health, and ergonomics. His research focuses on designing, validating, and disseminating human-centered, responsible, and affordable AI/ML solutions supported by federal, state, and private sector agencies. He serves on the editorial boards of the journals Construction Engineering and Management (ASCE) and Smart and Sustainable Built Environment (Emerald). His academic background includes a PhD in Civil Engineering from the University of Michigan, Ann Arbor, and a BS in Civil Engineering from Sharif University of Technology in Iran.

Research topics

• Computer Science
• Artificial Intelligence
• Engineering
• Computer network
• Construction engineering
• Programming language
• Operations research
• Transport engineering
• Medicine
• Data science
• Forensic engineering
• Geography

Selected publications

• The Influence of Information Source and Modality on Flood Risk Perception and Preparedness

  2025-12-07

  articleSenior author
  PublisherDOI

• Can LLMs assist in job interview preparation? Assessing the quality and effectiveness of LLM-generated feedback

  AHFE international · 2025-01-01

  article

  Large language models (LLMs) have demonstrated strong reasoning capabilities, making them potential candidates for generating formative feedback in learning contexts. This paper evaluates the ability of LLMs to provide formative feedback on interviewees' responses in a job interview task. Specifically, the degree of explanation in an interviewee’s response, a key communication skill, was used as the focal assessment criterion. Combinations of LLM models (i.e., GPT-3.5-Turbo, Gemini-1.5-Pro) with various chain-of-thought (CoT) prompting strategies, including task definition, domain knowledge, and contrastive prompting, are examined across multiple self-reported metrics of feedback quality effectiveness. Data was collected from 663 participants on Amazon Mechanical Turk using a between-subjects design with six experimental conditions, each corresponding to a combination of LLM model and prompting strategy. Results indicate that users perceived LLMs as having a moderate ability to provide formative feedback for job interviews, though the feedback was at times viewed as irrelevant or potentially harmful. The choice of LLM model and prompting strategy significantly influenced perceived feedback quality, with GPT-3.5-Turbo generally rated more favorably than Gemini-1.5-Pro. While stronger task performance occasionally aligned with higher user ratings, the relationship between performance and perception was not strictly linear. These findings are discussed in terms of design implications for enhancing the quality and effectiveness of LLM-generated feedback in interview training contexts.

  PublisherDOI

• User-Centered Design and Usability Evaluation of a Floodwater Depth Estimation Mobile Application

  AHFE international · 2025-01-01

  article1st authorCorresponding

  This study presents the user-centered design (UCD) and evaluation of a mobile application prototype, named Blupix mobile, which uses artificial intelligence (AI) and crowdsourcing to estimate the depth of floodwater in a user’s surroundings. Through a three-phase mixed methods technique based on UCD principles, the functionality, design, and usability of the app prototype are tested with a sample pool of U.S. participants. Results indicate a strong demand for location-specific, real-time alerts as well as community-generated content. The findings of this study aim to contribute to the expanding body of research on mobile disaster risk communication tools that incorporate community participation and engagement.

  PublisherDOI

• Factors influencing human trust in intelligent built environment systems

  AI and Ethics · 2025-08-14 · 2 citations

  articleOpen access1st authorCorresponding

  Abstract Artificial intelligence (AI) is rapidly integrating into infrastructure planning, design, construction, management, and operation. This encompasses the use of AI-powered, intelligent systems to process vast amounts of data and support human decision-making concerning urban development, architecture, transportation systems, housing, energy efficiency, and sustainability of the built environment. AI integration into the built environment has also introduced new challenges with respect to transparency, accountability, fairness, and reliability in machine decision-making, and led to concerns about algorithmic bias, privacy, and ethical deployment of technology. A key barrier to formalizing trust in AI lies in the absence of consistent definitions of the main constructs of trust, and their interplay in trust formation and calibration. Trust formation involves the initial establishment of confidence in AI systems, while calibration refers to the ongoing alignment of trust with system performance and user experiences. Understanding how these processes interact within a decision context is crucial for developing robust frameworks that enhance trust in AI. In the built environment domain, in addition to technical capabilities of the AI systems, it is also essential to consider social, ethical, and legal dimensions, to ensure that the outcomes serve the best interests of humanity. This paper provides an overview of trust in AI systems, followed by describing findings from the literature through the lens of several built environment decision-making scenarios with the goal of creating a common understanding to support future research on human trust in intelligent systems, and how it may influence the quality and timeliness of resulting decision outcomes.

  PublisherOA PDFDOI

• A Roadmap to the Next-Generation Technology-Enabled Learning-Centered Environments in AEC Education

  Journal of Civil Engineering Education · 2024-02-28 · 6 citations

  articleOpen access

  The architecture, engineering, and construction (AEC) education communities are increasingly facing challenges caused by social, technological, economic, environmental, and political changes. Addressing these issues requires AEC educators and practitioners to systematically rethink and reform many of their current practices. Anecdotal evidence in AEC education already exists with respect to pedagogical improvements made by individual technologies such as immersive computing, artificial intelligence, robotics, big data, cyberinfrastructure, and photogrammetry. However, an effective learning-centered environment is more complex than what any single technology can accomplish. In addition, the relationship between technology-intensive learning and digital inequity in AEC education remains, to the most extent, unclear. We envision the next-generation learning-centered environment for AEC education to be technology-intensive, interdisciplinary, industry-linked, and equitable. This paper aims to present a shared vision of the next-generation learning-centered environment for AEC education. To achieve this goal, two interrelated workshops were organized with the participation of different stakeholders, including researchers, educators, and professionals from multiple disciplines of architecture, engineering, construction, computer science, learning science, education, and social sciences. This paper is based on the combined outcomes of the two workshops, organized in four themes: (1) AEC curricula and industry practice, (2) technology and learning, (3) interdisciplinary education, and (4) digital inequity. This paper contributes to the body of knowledge by creating a pathway to timely reflect on new learning strategies, new technologies, and future industry and societal needs in AEC curricula, thus producing a more adaptive AEC workforce for the 21st century. The findings of this work can be adopted by educators to develop a roadmap for creating the shared vision of the next-generation learning-centered environment for AEC education.

  PublisherOA PDFDOI

• Rapid and Automated Vision-Based Post-Disaster Building Debris Estimation

  2024-01-25

  articleCorresponding

  Debris removal is one of the most expensive and challenging aspects of recovery from disasters. Erroneous debris estimation and delayed cleanup operation can block or slow down access to affected areas, thus putting public health at risk, and seriously hinder effective post-disaster search-and-rescue (SAR) and resource allocation. Existing disaster debris estimation techniques merely produce rough estimates of debris volume with significant errors. Recent research has demonstrated the value of low-cost unmanned aerial vehicles (UAVs) and artificial intelligence (AI) to rapidly collect information and assist in performing damage assessment of the building stock. However, the utility of AI models in debris estimation and classification is still underexplored. To this end, this study aims at enabling the measurement of debris volume and composition in residential structures from the outcome of AI-based damage assessment. Results from scaled experiments indicate that the proposed approach can provide high-fidelity estimation of disaster debris volume and composition.

  PublisherDOI

• Pedestrian Phone-Related Distracted Behavior Classification in Front-Facing Vehicle Cameras for Road User Safety

  2024-01-25

  articleSenior authorCorresponding

  Understanding distracted pedestrian behaviors is critical to road user safety and preventing traffic-related injuries. Front-facing vehicle cameras (a.k.a., dashcams) have increasingly become popular for documenting driving behavior and patterns. However, a relatively underexplored application of dashcam footage is to automatically identify distracted road users that may pose a threat to pedestrians. Detecting distracted behaviors in dashcam-captured imagery can enable drivers to take preventative measures and avoid potential traffic accidents. To this end, computer vision techniques powered by the prediction capability of artificial intelligence (AI) can be leveraged to identify pedestrians’ distracted behaviors when crossing streets and intersections. In this paper, pedestrians’ phone-related distracted behaviors are detected and classified in dashcam footage by leveraging convolutional neural networks (CNNs) assembled in the form of a two-stage detection and classification architecture. In particular, we propose to first detect pedestrians (stage 1) followed by classifying the most prevalent distracted behavior visible in each detected instance (stage 2). This technique has been developed on an in-house video dataset collected from urban intersections around a major university campus. Results indicate that the pedestrian detection model achieves 76% average precision (AP), and the classification of distracted behavior reaches 72% precision, 98% recall, and 83% F1-score.

  PublisherDOI

• A post-hurricane building debris estimation workflow enabled by uncertainty-aware AI and crowdsourcing

  International Journal of Disaster Risk Reduction · 2024-08-27 · 6 citations

  article
  PublisherDOI

• Capitalizing on strengths and minimizing weaknesses of veterans in civilian employment interviews: Perceptions of interviewers and veteran interviewees

  Military Psychology · 2024-05-23 · 2 citations

  articleOpen accessCorresponding

  = 93). Qualitative analysis of the focus group transcripts resulted in the emergence of two theme categories: (1) veteran interviewee strengths and (2) veteran interviewee weaknesses. This information guided the development of a 10-item survey that was completed by 93 veterans (Study 2). In its totality, the results (from both Study 1 and Study 2) indicated that communication of soft skills, confidence, and professionalism were perceived to be strengths that veterans displayed during civilian employment interviews, and conversely, the ineffective translation and communication of relevant technical skills acquired in the military, use of military jargon, and nervousness were considered to be weaknesses. Recommendations to capitalize on the strengths and mitigate the weaknesses are presented.

  PublisherOA PDFDOI

• Formalizing Trust in Artificial Intelligence for Built Environment Decision-Making

  AHFE international · 2024-01-01 · 2 citations

  article1st authorCorresponding

  While artificial intelligence (AI) has transformed the planning, design, construction, and operation of physical infrastructure and spaces, it has also raised concerns about algorithmic bias, data privacy, and ethical use in built environment decision-making. Addressing these issues is crucial for designing, developing, and deploying trustworthy AI systems that promote human safety, infrastructure security, and resource allocation. This paper reviews trust issues in AI through the lens of several built environment decision scenarios, e.g., weather prediction, disaster mitigation and response, urban sensing, and bridge health monitoring. It then outlines a framework to formalize trust, aiding researchers, policymakers, and practitioners in designing AI systems that serve societal interests.

  PublisherDOI

Recent grants

• Collaborative Research: Transforming Teaching of Structural Analysis through Mobile Augmented Reality

  NSF · $30k · 2018–2021

• Collaborative Research: Strategies for Learning: Augmented Reality and Collaborative Problem-Solving for Building Sciences

  NSF · $37 · 2017–2018

• Real-Time Feedback-Enabled Simulation Modeling of Dynamic Construction Processes

  NSF · $134k · 2017–2019

• Real-Time Feedback-Enabled Simulation Modeling of Dynamic Construction Processes

  NSF · $275k · 2016–2018

Frequent coauthors

• Reza Akhavian

  San Diego State University

  50 shared

• Vineet R. Kamat

  26 shared

• Semiha Ergan

  San Diego State University

  25 shared

• Fei Dai

  20 shared

• Jing Du

  University of Florida

  20 shared

• Nipun D. Nath

  17 shared

• Theodora Chaspari

  University of Colorado Boulder

  17 shared

• Yalong Pi

  Texas A&M University

  10 shared

Education

• Ph.D.

  University of Michigan

  2008

• M.S.

  University of Michigan

  2005

Awards & honors

• Faculty Fellow, University of Colorado Boulder, Research & I…
• Plenary Speaker, Annual Conference of the International Asso…
• "Editor's Choice" recognition, ASCE Journal of Engineering M…
• Top Downloaded Article, Journal of Computer-Aided Civil and…
• Second Place Award, The Inaugural Best Data Award Competitio…