Drone Inspections for Solar Farms | Saving Time, Boosting Output, and Cutting Risk

Infrastructure doesn’t fail all at once. It fails slowly, quietly, and often out of sight. Traditional inspections are expensive, slow, and risky. Whether it’s bridges, towers, culverts, or rooftops, getting accurate data used to mean shutting things down, climbing into dangerous spots, or relying on outdated reports.

In this episode of The Drone Advantage Podcast, Rob breaks down how drone inspections are helping cities, utilities, engineers, and property managers get faster, safer, and more accurate infrastructure data, without the cost, downtime, or liability of traditional methods.

You’ll learn how drones are being used to inspect:
– Bridges, rooftops, and elevated structures
– Stormwater systems, culverts, and retention ponds
– Communication towers and substations
– Commercial buildings and hard-to-access sites

We also demonstrate how drone imaging is used to create digital twins. These are realistic 3D site models that allow engineers and asset managers to track changes over time, detect damage early, and collaborate remotely using accurate, repeatable visuals.

We’ll talk about the ROI of drone-based inspections, how repeatable flight plans enable side-by-side comparisons, and how organizations are using this data to stay compliant, reduce maintenance costs, and improve asset oversight.

🎧 Hosted by Rob, FAA Part 107-certified drone pilot and owner of Blue Nose Aerial Imaging of Tampa Bay.

📅 Available on YouTube, Buzzsprout, Spotify, Apple Podcasts, and more!

📍 Follow for More Drone Industry Insights:
🔹 Buzzsprout: https://thedroneadvantage.buzzsprout.com
🔹 Website: https://get.bluenoseaerial.com/tampa/
🔹 LinkedIn: https://www.linkedin.com/company/blue-nose-aerial-imaging-of-southwest-florida/
🔹 Instagram: https://www.instagram.com/bluenosetampa/
📺 Watch full episodes on YouTube: https://www.youtube.com/@bluenosetampa

#DroneInspections #InfrastructureMonitoring #DigitalTwins #AerialData #MunicipalDrones #Stormwater #BridgeInspection #FAA107 #DroneAdvantagePodcast #AerialIntelligence #AssetManagement #UtilityInspections #DroneTechnology

Transcript

Solar systems don’t fail all at once. They slip. They go down one panel at a time. A combiner box shorts out, and if nobody’s watching, the losses, the expense, just adds up—day after day after day. That loss compounds fast. And that’s why thermal drone inspections aren’t just nice to have anymore. They are the only scalable way to keep your solar assets honest.

Today we’re breaking down how aerial thermography is saving clients real money, cutting repair times, giving operators hard proof of what’s failing, where it’s failing, how bad it’s failing. Whether you manage these assets or you fix them or you inspect them, this is one of the most powerful tools in your toolkit.

Have you ever wondered how aerial thermal imaging really works and how it’s changing the game for solar farms, but not just solar farms—also rooftop inspections, commercial energy management? If you’ve ever wondered any of this, this episode is for you. I’m going to walk you through how thermal inspections are performed, the kind of data we get from them, and most importantly how that data translates into real world decision‑making.

And just as a sneak peek: this is the last episode of Season 1. I’ll get into a little bit about Season 2 and what that’s going to entail at the end of this episode.

I’m Rob. I’m a commercial airline pilot, FAA‑certified drone pilot, and the owner of the Tampa Bay franchise of Blue Nose Aerial Imaging. You’ve tuned in to the Drone Advantage podcast where in Season 1 we’ve been breaking down how aerial intelligence has been changing the landscape of businesses all across the world and how you can use drone data to work smarter. Let’s get into it.

Let’s get into the basics. How do drone-based solar inspections even work?

Well, most commercial inspections use a thermal imaging drone like this one—the DJI Mavic 3 Enterprise. By the way, does anybody out there name their drones? If you have drones, do you name them? Leave that down in the comments, because all of my drones have names, and you’ll see on this one here this is my DJI Mavic 3 Thermal, and it is appropriately named Heatseeker. Anyway, fun fact.

So we go out to the solar farms and we’re doing these inspections. What is it we’re inspecting anyway? We’re taking this thermal camera on this drone and we’re actually looking at the temperature of whatever we’re looking at—in this case, a solar farm. We’re looking at the temperature of the solar panels and we’re getting that radiometric data.

What does radiometric mean? That means when the photo is taken—when the thermal image is taken—it’s stored as an R‑JPEG. “R” stands for radiometric. That means every pixel in that photo has a temperature assigned to it. So when we just look at the photo, all we’re going to see is blotches of colors, right, depending on what kind of color palette we’re looking at. I’m going to show you some a little bit later in this episode. So if you are listening on Apple Podcast or Spotify or whatever your favorite audio‑only podcast platform is, I highly recommend that you jump over to YouTube, subscribe to the YouTube channel—this podcast is on video—and I am going to be showing photos and maybe even a video of some solar farms, and I’m also going to jump into a roof inspection as well. Even though this episode is about solar inspections, I’m going to jump into a roof inspection so that I can show you a report. And it’s going to be a real report that I delivered to a client a couple of months ago.

So we take photos. Those photos are stored with radiometric data—meaning that each pixel has a temperature assigned to it—and then when we look at that photo, all we see are colors like I said, depending on the palette. But there is software that can then read the temperature data off of those pixels. Since I use a DJI platform, the software I use is the DJI Thermal Analysis Tool. It works great, and it is the software that I use to produce the report that I’m going to show you later.

People say that you can’t actually measure the actual temperatures of things with these thermal cameras. And I’m going to say that’s not true. I’m going to say it’s true most of the time, but it’s not true as long as you have your settings correct. If you have set your emissivity setting and your distance setting correctly, then you can get accurate temperatures as to what you’re looking at—or at least you can get temperatures that are accurate enough for you to make the decisions you need to make. And that’s all we’re looking for. We’re just looking for data to help us make business decisions.

So if we look at, MSI for example, every thermal camera—at least any thermal camera I’ve ever used—has a setting where you set the emissivity. What emissivity really is is just a scale from 0 to 1 as to how efficient a particular material is at reflecting heat. Zero being it absorbs all the heat, and one being it reflects 100% of the heat. Well, almost nothing reflects 100% of the heat. So generally, the emissivity of a particular material is going to be somewhere between 0 and 1, with some of the highest reflectance materials being 0.95, and that’s what we use for solar panels. It’s .95. But also I use .95 for some other things, and I may lower it—but you can look these up in tables. You look at the material that you’re imaging, find it in an emissivity table, get the number, put that into the camera settings—and then the camera will use that in its algorithmic calculation to tell you what the actual temperature of what you’re looking at is.

However, there’s one other setting that needs to be correct, and that is your distance from the object. Because as we get further and further and further from the object, less and less of that temperature energy—or that heat energy—reaches the camera sensor. So it needs to know how far it is from the object so that it can apply that calculation and give you an accurate temperature.

What’s nice about the DJI Thermal Analysis Tool is that even if you set it incorrectly in your drone, you can change it in the thermal analysis tool afterward. Another thing that’s nice is, whatever color palette you took your photos in in the drone, you can change that in the Thermal Analysis Tool. If you took your photos in white-hot but you want to look at them in iron grid, you can just change it in the tool without having to go reshoot the photos. And another thing DJI Thermal Analysis Tool does is it allows you to set the temperature range of the color palettes. On the controller itself, you can set a couple of temperature ranges, but you can’t set the upper and lower temperature ranges in a fine-grained way like you can in the Thermal Analysis Tool.

When I show you the inspection report of a thermal roof inspection I did—it’s not solar panels, it’s just a roof inspection looking for water leaks—you’ll be able to see on the side I narrowed the scale down to the temperatures I wanted to report to the customer. But since this episode is about solar, let’s go back to solar farms and stay there a bit.

When I go out and shoot a solar farm, it’s very important I do so under the right weather conditions. This is what separates the amateur pilot and the pro pilot: really understanding what you’re taking photos of, and the conditions in which those photos are going to be accurate and usable data. For example, when I go out and shoot a solar field, generally the customer will have a minimum irradiance spec they want the photographs taken at—and generally that number is around 600. It can be a little lower under certain circumstances, but irradiance of 600 or more produces the best, most accurate data. Why is that? Because at 600 irradiance or more, the solar panels are under full solar load—or at least increased solar load—so any problems or anomalies in the solar panels are going to show up, they’re going to be exaggerated. Versus if we went out and the irradiance was only 300, for example, those panels aren’t under a lot of load, and whatever might be getting hot on them may not be getting hot because there’s not enough current to get it hot. I’ll show you some photos I took for an inspection of connectors underneath the solar panels, not on the panels themselves.

Another thing that separates the pros from amateurs is how to measure irradiance. What equipment? Obviously you have an irradiance meter—but how do you use it? You can take an irradiance meter and wave it around the sky and get different numbers everywhere. So how exactly am I using it? And when am I shooting—time of day, what are the sky conditions like? Another thing people don’t think about: how windy is it? Some customers will have a spec that if wind is more than 15 miles an hour, you just don’t do the inspection. Why? Because as wind speed increases, hot things get cooled down and that might hide problems.

Top conditions: A1—solar noon, irradiance 600+. No clouds. No wind. That’s perfect. Then you start from perfect and figure out what’s acceptable. I can go down from my perfect 900 irradiance to the client’s minimum 600. Maybe that allows some fluffy clouds in the sky or me not being exactly at solar noon. But under no circumstances am I going to fly 2 hours before sunset or after sunrise—they just don’t get enough solar energy to show anomalies.

Knowing how to capture data is just as important. Like I said in the agriculture episode, there’s no magic drone that automatically collects data and gives answers. We have to know how to take raw data and apply context to make good business decisions.

What are we looking for when doing solar inspections? Temperature deltas—differences. Whatever temperature a panel is, it should be consistent across its surface. If we see hot spots or cold spots, that tells us something. My job isn’t to diagnose; my job is to point out problems. These drones aren’t replacing boots on the ground. They’re not eliminating humans—they’re directing them. I find a problem, send a tech, they verify and fix.

What kind of problems? Dead strings, dead cells. In solar farms that span hundreds of acres, just one dead cell can knock off a whole panel or string. An engineer trying to find every dead cell manually in a 400‑acre farm? Impossible. Drones are the only scalable method. Aircraft used to do it, but they can’t get low or slow enough—and their thermal resolution can’t detect a dead cell. You need high resolution. Drones can do that.

Let’s talk who needs solar inspections and why they keep writing the checks. In construction, agriculture, infrastructure, solar is no exception: companies aren’t increasing expenses—they’re decreasing long-term costs. With solar, if a panel, a string, or a cell goes down, it’s not producing at the calculated efficiency. That means less electricity sold, which means lost revenue daily. If a bad string goes unnoticed for a month or two, the losses are way higher than paying a drone service provider and an engineer to fix it.

We’re basically helping a power company keep their commodity—electricity—flowing. If they don’t produce, they don’t sell. That’s why they keep writing checks: they’re protecting revenue by catching faults early.

Next, I want to dive into an inspection I did—weeks long—for a huge power company looking for hot connectors. Here’s what I want to do: show you what we’re looking at.

These images are the underside of a panel, and that connector right there is what we’re inspecting. We’re looking for hot connectors. The customer spec was white-hot palette, and in the thermal photo, that connector is white. That shows it’s very hot.

If I put this into the DJI Thermal Analysis Tool, I can tell you we were using Celsius, and that connector was around 75 °C. I had that on screen when taking the photo—it doesn’t show up on the image itself, but since it’s an R-JPEG, the temperature data is stored in each pixel. I could measure it later at around 73–75 °C.

Here’s another example: a photo of three connectors. Which one’s bad? Maybe all of them. The photos are geotagged; I provide a KML for the client so they can drive directly to each connector. Then they scan it with a handheld thermal scanner, verify the issue, and replace it.

Here’s a side-by-side: the RGB photo of the connector and the thermal image. All of these were above 60 °C because the spec was 30 °C above ambient. Ambient that day ranged from 28 to 32 °C, so the connectors were well above 60 °C.

That week I did about 11 batteries a day. Every battery change, I’d check outside temperature, adjust my controller’s thermal alarm settings. If ambient was 31 °C, I set alarm to 61 °C. If ambient dropped to 29 °C, I brought it down to 59 °C. Almost all photos were taken with ambient ≥30 °C, so above 60 °C.

I didn’t show these in Thermal Analysis Tool because it doesn’t work on Mac—and I was on Mac. But I have it on PC to build the report. I want to show you this thermal roof inspection report: it’s several pages with photos flagged in boxes. In the Thermal Analysis Tool you can see the temperature range I highlighted—based on 110° to 187°F, for example. I set the range to highlight the actionable zone and everything else is background.

What the tool does: drag the palette range so deep purple is your lower threshold, deep red your upper. Anything outside is noise. In the selected box, the tool points out the hottest and coldest pixel. Actionable data.

Here’s why it matters. I drew a square, the tool tells me min is 126°F, max 184°F, average whatever. Those are actionable. It's not noise. And yes, while emissivity and distance settings aren’t perfect to the tenth of a degree, this data is accurate enough to act on.

But context? Critical. Without context, photos mean nothing. You need to know the time of day, outside temperature, solar position. Why? Because thermal imaging detects water leaks by contrasting thermal inertia. Water heats and cools slower than roofing material. In the morning, trapped water appears cool compared to the roof. In the evening, it holds heat and appears hot compared to the cooling roof. Without context, you’d misinterpret.

In this sample photo: air conditioning units show expected cold. Some areas show roof leakage. If I’d taken it in the evening, I might’ve thought opposite. That’s why context is everything.

That roof was a mess—four or five flagged boxes. Each box had its own measurements listed down in the report. The tool lets me change color palettes before running and exporting the report. Once I pick white-hot or iron bow, it renders the report accordingly.

You might ask: why not just use a handheld thermal scanner on the roof? Why hire a drone service provider? You still need deep thermal to detect leaks, but with handheld scanners you’re climbing on dangerous roofs. That particular roof was 9,000 sq ft and I scanned it in 10 minutes. A contractor would take much longer. Plus risk of falls, heat, liability. Insurance loves it when you use drones because it reduces injury claims.

On top of that, I delivered a PDF report with geo-tagged, annotated, temperature-data-backed images. The contractor loved it. He’d never seen something so useful and accessible.

Wrapping up the solar episode…
Okay, I dove more into roof inspections than I planned, but it all tied together. If you’re a roofing contractor or a homeowner, thermal roof inspections are key—especially in Florida before hurricane season. You want baseline data before storms hit. That data acts as evidence during insurance claims. Helps prove pre-existing conditions, avoid litigation, and get paid fast.

If you work in solar—commercial or residential—start using thermal drone imaging. Analyze system integrity. Make sure everything’s producing. Identify faults before you lose money daily.

Season 1 wrap-up—I said at the start this was the final episode of Season 1. Season 1 was directed at business decision‑makers: agronomists, superintendents, project managers, solar farm owners. We discussed the ROI of drone data. If you haven’t seen previous episodes, subscribe on YouTube or Spotify/Apple/Buzzsprout, then go back and watch. Even if you’re in construction, agriculture, or infrastructure, the data can boost ROI.

Season 2 preview—we’re shifting gears. Season 2 is for drone pilots—you—the person planning missions, choosing equipment, working software. We’ll cover mission planning, gear selection, RTK, PPK, ground control points, DJI Pilot 2 settings for multispectral, thermal, Mavic 3 Enterprise, software workflows in Pix4Dmatic, Pix4Dcloud, DJI Terra. More episodes, each with clear takeaways to improve your next mission.

If that sounds like where you want to be, hit subscribe on YouTube, Spotify, Apple, Buzzsprout, wherever you listen. Season 2 is coming for drone professionals who want to fly smarter.

Until then, thanks for listening to The Drone Advantage Podcast—keeping business decisions under 400 ft.