Postdoctoral Researcher – Buildings
University of California – Los Angeles | Civil and Environmental Engineering

Hesam Soleimani-Babakamali (he/him) is a postdoctoral researcher from Iran based at the University of California – Los Angeles. He has a master’s degree in civil engineering from Sharif (Aryamehr) University of Technology. Hesam earned another master’s degree in computer science at Virginia Tech, where he also graduated with a PhD in civil and structural engineering in 2022. A native of Tehran, Hesam now resides in Blacksburg, Virginia.

Journey to the Hub

I joined CHEER in November 2022, so I think I’m one of the longest-standing researchers in the Hub. Joining the team was a natural progression from the work I was already doing. Three years ago, the application of deep learning methods, an emerging paradigm in artificial intelligence, to civil and environmental engineering problems was still very young. The field held immense untapped potential, but individuals with strong foundations in both civil engineering and computer science were far less common than they are today. Given the scale and complexity of the CHEER project, this interdisciplinary expertise was crucial. As a result, the onboarding process was both smooth and swift, and the role provided a unique opportunity to apply my data science skills to a novel and impactful domain that deals with the well-being and prosperity of people.

Research is my hobby, a passion of mine that I may not have dreamed of as a child, but one that has become a core part of who I am. I genuinely enjoy conducting research. I’ve worked independently on a range of topics in disaster science: pioneering the use of generative AI in structural health monitoring, applying AI in earthquake engineering, and developing post-hazard rapid damage reconnaissance frameworks for the 2023 Türkiye earthquake and the Lahaina wildfire. None of these efforts were part of a job, scholarship, or internship; I pursued them purely out of personal interest and conviction. These projects have been rewarding because they allowed me to work freely, driven by curiosity and a desire to contribute meaningfully.

 

CHEER Roles and Responsibilities

I’m a member of the CHEER’s buildings thrust, where – aside from the various structural and hazard loss tools we develop – working with big data is an integral part of our research. “Big data” is a term that refers to data sets that are too large or complex to be dealt with by traditional data-processing software. In our thrust’s case, this information includes statistics on millions of acres of land and buildings as well as large-scale socio-economic data covering the Hub’s three case study areas. To handle these vast datasets meaningfully and efficiently, we leverage the most recent advancements in artificial intelligence, such as computer vision, regression, and data imputation. Additionally, the resulting structured databases and reproducible code – all supported by detailed documentation – can provide valuable resources for the broader research community.

My research focuses on big data and AI applications. My efforts on the buildings thrust are two-fold: establishing large datasets that capture both basic structural characteristics (e.g., number of floors or whether a structure is wooden or steel-based) and socio-economic features (e.g., asset value or estimated household population) of designated structures by merging public data from various sources (e.g., tax data, parcel data). I use AI models to enhance these datasets by filling in any missing records with hypothetical data. Filling in those “blanks” allows our team to draw more accurate conclusions and infer new attributes of the structures that are not revealed in public data. One example that might be missing in public sources is whether a structure has a gable, hip, or other roof types, a key characteristic that significantly influences structural performance during hurricanes and related hazards.

Across all of my work, I am deeply committed to addressing the challenge of generalizability — ensuring that models maintain robust performance when applied to new and unseen data. For instance, if a computer vision model is trained to estimate the percentage of openings on building facades, such as doors and windows – some of the most vulnerable points through which hurricane hazards can penetrate a structure – does it maintain its performance across images with different lighting or ambience? Can it remain statistically unbiased when applied to images from various counties within the study region? Developing more robust and less biased data would enable the building thrust to generate more reliable outputs for downstream CHEER modules that ultimately inform policy-making decisions.

 

Working with the CHEER Team

The Hub is a massive project, and with large projects come many people. CHEER is a truly wonderful collection of researchers. While most of my interactions have been with the buildings team, the multi- and interdisciplinary nature of the project has given me the pleasure of working with people across various disciplines. With all of our team members’ expertise comes a wealth and breadth of knowledge. CHEER has a lot of moving parts; having this broad understanding of the work our team is doing at large helps me connect the dots between the Hub’s various research activities and endeavors.

How are hazards simulated? How do they interact with certain structures? How are people within those structures impacted by the hazards? How do all of these elements merge to inform policy? All this broad knowledge is part of what I’ve learned from colleagues, made possible only through the collaborative research community that CHEER has built.

 

Hub Highlights

As I’ve mentioned, some of the research I’ve pursued in the past was done independently, driven by my passion for the sake of scientific discovery. These projects have been some of my favorites. That being said, one of the unique strengths of the Hub has been its support for me to continue this kind of independent research.

Given CHEER’s broad scope, new research ideas also often emerge from many directions. Even when these ideas fall outside of the Hub’s research goals, I’ve frequently recognized their potential to contribute meaningfully—either as complementary efforts or as parallel initiatives to CHEER’s core work. For instance, I’ve had the opportunity to conduct research on rapid post-hazard damage assessment (think wild and urban fires and earthquakes) using remote sensing data and the development of scalable Structural Health Monitoring systems. In doing so, I drew on new expertise and techniques developed through my work at CHEER.

Additionally, CHEER has a SEED funding program to support initiatives that don’t fall directly under its focus areas. It’s rare to find a postdoc position that encourages such latitude, and I’m glad to be a part of a team where I can pursue mutual research interests that intrigue me, challenge me, and are meaningful.

 

Looking Forward

My experience in CHEER has helped me grow in every aspect – professionally and personally. Working in such a large and interdisciplinary group has improved my ability to collaborate and communicate. I’ve expanded my expertise, worked with types of data I had never encountered before, and developed new skills. Whether I go on to pursue a traditional 9-5 job is another matter, but the growth I’ve experienced through CHEER is, in many ways, all I could ask for. It may even be more than enough to embark on a new career path or build something of my own.

 

Spotlight Trivia

In addition to conducting research, collecting and reading comics – specifically, Japanese-style comics, or Manga – is a passion and hobby of mine. Over the past few years, I’ve developed many of the skills and strengths I need to seriously pursue a career as a comic artist and author alongside my research. The dark-fantasy story I have in mind is going to be very long, but I’ve already mapped out every single detail for the first 50 chapters. All of the characters have been fully developed and designed, and my digital art skills have steadily improved. I’ve also built strong relationships with many supportive collaborators. All that’s left is to bring these elements together and begin publishing substantial chunks of chapters. My plan for the future is to continue embracing the freedom I have in my research while adding a grain of Manga to the journey.