Drones have long been associated with building inspections, as drones don’t need ladders — and they can see the tops of roofs or high places without sending a person up there. But more drones can mean more data to analyze. And more data to analyze just creates more work, which isn’t necessarily a good thing. That’s why AI software is becoming critically important in all facets of the drone industry — but especially with building inspections.

Lamarr.AI is one of those companies using AI in tandem with aerial thermal imaging to change how we diagnose the health of buildings. Now that Lamarr.AI has completed multiple, successful municipal pilots including one with the City of Detroit, I chatted with their team to find out more about their AI software and what the next generation of smart infrastructure management could look like.

How Lamarr.AI works to make building inspections faster

Using drones equipped with visible-range and thermal cameras, Lamarr.AI captures detailed data about a building’s exterior. But flying the drone is not enough to really make building inspections meaninfully faster.

The real magic happens after landing. That’s when the data is fed into Lamarr’s proprietary AI software, which generates insights such as 3D models, energy simulations and prioritized retrofit recommendations.

“We deliver 6 to 10 reports per week depending on the tier,” said Tarek Rakha, CEO and co-founder of Lamarr.AI. “Our Audit-level reports take about a week to complete while Detect-level reports are typically turned around in under 3 days. This output already provides a meaningful impact for our clients but as we continue to refine and automate our AI pipeline, we expect our capacity to scale significantly by the end of this year.”

In a recent Detroit deployment, Lamarr.AI identified over 460 thermal deficiencies across several city-owned buildings. Rakha says such insights would traditionally take weeks of manual inspection.

Related read: Detroit: from Motor City to Drone City?

And after identifying those deficiencies, it was able to make recommended fixes, ranging from insulation upgrades to specific window replacements and weatherization at the curtainwall. Then, the software is able to prioritize it all by severity.

What it’s like working with Lamarr.AI and its AI software

Lamarr.AI’s business model is flexible. The company operates as both a full-service diagnostics partner and as a software platform for others to use.

“We do both,” said Norhan Bayomi, Chief Product Officer. “We’re an end-to-end platform for building envelope diagnostics and solutions working directly with large portfolio owners and operators. At the same time we license our platform to owners who have their own drones or third-party drone service providers. This dual approach allows us to maximize our impact across the industry.”

For drone service providers, this presents a unique opportunity to up-level their offerings — not just delivering imagery but delivering real building intelligence.

Why AI software is the critical component

What sets Lamarr.AI apart isn’t just automation — it’s intelligence. The platform doesn’t just detect issues; it evaluates their importance using expert-informed AI.

“Our system replicates a building scientist’s decision-making by fusing information from both visible-range and thermal images and analyzing shape, intensity and location,” Rakha said. “It uses proprietary neural network models trained on expert-annotated data and developed for this domain-specific challenge plus rule-based logic to classify severity — flagging urgent issues such as water intrusion as high priority and categorizing minor inefficiencies separately.”

Behind the scenes, Lamarr’s models were built using a robust dataset from a U.S. Department of Energy-funded project called AirBEM, which was developed through research partnerships with MIT, Georgia Tech and Syracuse University.

“Our detection models were trained on a large expertly-annotated dataset of thermal and RGB images covering key envelope issues like insulation defects, air leakage and moisture intrusion,” said Senem Velipasalar, CTO of Lamarr.AI. “We continue to expand our dataset with every project that allows us to use their data, ensuring our models stay accurate and cutting-edge.”

As more cities pursue climate-conscious infrastructure strategies, Lamarr.AI sees its role growing beyond building audits and into the fabric of urban operations.

“Our vision is to be the digital backbone of citywide building exterior diagnostics,” said John Fernandez, Lamarr.AI’s Chief Strategy Officer. “From our pilot in Detroit to ongoing projects across the U.S., Canada and the UAE, we’re showcasing that our technology can seamlessly integrate with municipal asset management systems in a way that was not realisable before. By providing real-time insights and predictive maintenance tools, we empower cities to make cost-effective data-driven decisions that enhance energy efficiency, resilience and sustainability at scale.”

Humans still have jobs — for now

While Lamarr.AI’s software is highly automated, the company still maintains human quality assurance as part of the workflow.

“At this time we require human-in-the-loop validation to ensure every report meets our high quality standards,” Bayomi said. “This step ensures that every dataset and recommendation is thoroughly vetted by our team of experts. As we continue developing our AI, the goal is to minimize the need for human oversight while empowering our building and computer scientists to expand our inspection capabilities beyond what’s currently possible.”

This article was sponsored by….YOU! For more industry insights and interviews with the players who keep pushing drone tech forward, please donate to The Drone Girl! Use the button below or go through my PayPal to donate. Thank you!

The post How AI software is transforming building inspections with drones appeared first on The Drone Girl.