Utilizing AI in Building Performance with Dr. Tarek Coehn’s Lamar.ai

The Office of Sustainability had the chance to speak with Tarek Rakha, Associate Professor at Georgia Tech’s School of Architecture and Director of the High Performance Building Lab. He is currently on leave as a Regents Innovator to commercialize his research through Lamar.ai, a company he co-founded. Lamar.ai leverages drone technology and AI to revolutionize building diagnostics, offering a faster, cheaper, and more accurate alternative to traditional inspection methods.d Resilience Specialist, in the Office of Sustainability

Can you please tell us a little bit about yourself and your background and how you ended up in Atlanta?

I’m originally from Cairo, Egypt, where I trained as an architect. I completed my bachelor’s and master’s degrees at Cairo University and later pursued a Ph.D. in Building Technology at MIT, after initially starting at Harvard’s Graduate School of Design. My doctoral work focused on urban systems — including thermal comfort, mobility, and how these factors influence decisions in the built environment. I began my academic career at Syracuse University, where I first explored the idea of using drones for thermal mapping in 2015. In 2019, I joined Georgia Tech, where I led the High Performance Building Lab and received a major research award from the U.S. Department of Energy called AIRBEM (Aerial Intelligence for Retrofit Building Energy Modeling). That project significantly advanced our technology and ultimately led me to co-found Lamar.ai, where I now serve as CEO while on leave as a Regents Innovator.

Can you tell us more about the drones and your company?

At Lamar.ai, we describe what we do as being the ‘MRI for buildings.’ We use drones to capture both visible-light and infrared images of building exteriors. These images are analyzed with computer vision algorithms we’ve developed over the past decade to detect and diagnose issues such as air infiltration or exfiltration, thermal bridging, water intrusion, and physical damage like cracks. All of this data is mapped onto 3D models, creating a comprehensive visual representation of a building’s condition. Our platform offers three core services: Lamar Detect, which identifies anomalies quickly and cost-effectively; Lamar Diagnose, which provides work orders and solutions; and Lamar Audit, which calculates return on investment and energy savings. Our system is 5–10 times cheaper, up to 10 times faster, and about 50% more accurate than traditional methods — and much safer because no one needs to physically climb buildings. We’ve deployed this technology across North America, the UK, and the UAE, supporting both existing buildings and new construction projects.

How do you envision projects like this working at Georgia Tech?

Georgia Tech’s campus is large and complex, with many buildings that have deferred maintenance needs, especially related to energy performance. Our platform could provide a scalable solution by conducting campus-wide assessments — for example, inspecting 25–30% of buildings each year. This would allow Georgia Tech to prioritize which buildings need immediate attention, whether that’s weatherization, roof replacement, or targeted repairs. We can even detect early signs of mold or HVAC system issues and verify contractor work after repairs are completed. Importantly, our approach is also far more cost-effective: a traditional building enclosure investigation might cost $25,000, but we can conduct the same assessment for around $3,000. By leveraging this technology, Georgia Tech could significantly reduce costs, proactively plan capital investments, and improve building performance across campus.

Is there an opportunity to involve students and faculty in this work?

Absolutely. There are multiple ways to integrate Lamar.ai into educational programs. Students could participate in data collection by learning to operate drones, which would be especially relevant for aerospace engineering courses. On the analytics side, computer science and electrical engineering students could work on refining our AI models or annotating data. Architecture and design students could use the insights from our platform as part of design studios or competitions, similar to what we’re doing with the University at Buffalo’s ‘Resilient Campus’ competition, which focuses on sustainability-driven design proposals. These opportunities not only provide hands-on experience but also help bridge research, education, and real-world impact.

What are the broader sustainability benefits of this technology?

Our work has direct implications for sustainability. By detecting and addressing building inefficiencies, we can significantly reduce energy use, improve occupant comfort, and enhance structural safety. All of this translates into measurable reductions in greenhouse gas emissions, which aligns closely with Georgia Tech’s Climate Action Plan. Additionally, because our decisions are based on real data rather than assumptions, they are more cost-effective and targeted, leading to better resource allocation and a higher return on investment. Ultimately, this technology supports not just sustainability goals but also operational efficiency, occupant health, and long-term building resilience.

Do you have any final thoughts to share?

We’re at a pivotal moment in sustainability and climate action. The challenges we face globally are significant, but the technology now available to us makes it possible to address them more effectively and affordably than ever before. By leveraging data-driven tools like Lamar.ai, we can make informed decisions that not only improve building safety and performance but also reduce environmental impact. My hope is that Georgia Tech — where this idea was born — will fully embrace this technology as part of its sustainability strategy, setting an example for institutions around the world.