After running a thermal drone scan, Jeff and his team at Structura View identified 121 locations on this massive industrial roof that are suspected to be wet. But thermal data alone isn’t enough—verification is key.
Next week, they’ll be on-site to confirm moisture levels, analyze seam conditions, and determine if material differences or uplift are contributing to these anomalies. Every thermal signature tells a story—it’s our job as consultants to understand why it’s happening and how to fix it before it gets worse.
Both Jeff Carrillo from Structura View and Marvin Rosario from Airweb Digital
, provide not just the flight but the full-picture insights that help utilities and infrastructure teams make smarter decisions.
TRANSCRIPT
Jeff:
We just did an inspection, a thermal roof scan with a drone on a 1. 2 ish million square foot industrial building. So essentially huge 36 acre flat roof, all white. So it’s a single ply TPO membrane, excuse me, two year old building. I’m in central Florida. We obviously just had a couple of hurricanes hit us this last season.
The property owner also had a lightning strike on the building, so they knew they had some issues, but they didn’t understand the extent of it. And they had some inspections done on the building that came back saying that there was no additional damages found. They had hired us to do a thermal scan as essentially a little bit of a fact check. And we can go into more of the science behind thermal imaging and how it can be a good resource for this type of stuff. But we ended up finding 121 different locations on the roof that we suspect to be wet.
And we are working on, I think next week, we’re probably going to be going out there to do some verification to actually make sure that they are wet and see if we can correlate causation.
Is there an opening? Are the seams failing due to uplift? What, what’s going on? Or is it not wet and we have differences in material? There’s usually a reason as to why there’s a thermal anomaly that shows up. We just have to understand why, right?