Friday, 6:47 PM. I was shutting down my laptop when the phone rang.
A facilities manager from a food-processing plant in Wisconsin was calling. Their chiller had tripped twice in nine days. Two contractors gave two different diagnoses, and the second shutdown had caused somewhere around $11,000 in lost production time. They wanted the highest-resolution consumer thermal camera we sold, overnight.
I get some version of that call at least once a month. Usually on a Friday.
Resolution is real. But it is not a plan.
Spec questions dominate this industry. People search for the FLIR highest resolution consumer thermal camera or look up FLIR vs. Fluke threads, and turn a purchasing decision into a numbers contest. Pixel counts feel safe. They are comparable even when you don't yet understand why one camera costs five times more than another.
When you spend your days triaging rush orders, you notice something about those calls: "highest resolution" means a different thing depending on who is saying it.
- Sometimes the caller needs to spot a corroded connection from 40 feet away.
- Sometimes they have to document findings for an insurance or safety audit.
- Sometimes they have no specific requirement at all. "Highest resolution" just sounds like the responsible choice.
That last group is larger than you'd think. And I've helped too many people in that group buy the wrong tool.
Here's the part that took me a while to learn. It took about four years and 150-plus on-site callouts to understand it fully: thermal resolution and temperature-measurement quality are not the same thing. A 640×512 sensor doesn't "measure better" than a 160×120 sensor. It gathers more thermal data points across the scene. That matters when you're trying to see a small object from a long distance, or separate two components that sit close together.
But the accuracy of a temperature reading comes from the detector's sensitivity, its calibration, and how the camera processes radiation. A high-pixel-count imager with a wobbly calibration can drift more in one work shift than an industrial unit with a quarter of the pixels drifts in a year.
People assume more pixels means more truth. It doesn't. Aiming a high-resolution sensor at the wrong target just produces a sharper picture of a missed problem.
Oh, and one more thing about that search phrase. As of January 2025, consumer-class FLIR models that attach to a smartphone top out at a 160×120 sensor, per the FLIR product pages. That isn't an oversight. A larger detector changes the size of the lens, the power draw, and the price point, so it stops being a pocket accessory at some stage. The better question is not "which has the most pixels in the consumer category?" It's "do I need consumer portability or professional measurement capability?"
The same mistake shows up in other industries
A colleague who sells operating microscopes tells me his customers run into the exact same pattern. A dentist will ask whether they should choose Zeiss vs Global dental microscope, and the first thing they compare is maximum magnification.
His answer, repeatedly, is that both lines have more than enough magnification for detailed clinical work. The differences that matter are working distance, depth of field, and how the scope handles posture, because a dentist stays in that eyepiece for hours. Magnification is just the easiest number to put side by side.
Thermal cameras behave the same way. Resolution is a number you can put side by side. Field of view, lens options, measurement accuracy, frame rate, and reporting workflow are harder to compare, so they get pushed aside. But those are the specifications that actually decide whether the tool fits.
Start with the tools you already carry
Let's bring this down to a realistic example, because talk of sensors can stay abstract for too long in this industry.
Take an HVAC service technician who shows up with a Fluke 116 HVAC multimeter. That meter is ideal for residential and light-commercial HVAC because it reads microamps for flame sensors and includes a built-in thermometer. Its sibling, the Fluke 117 multimeter, is the go-to for commercial electricians because it adds non-contact voltage detection and True RMS measurements.
Both are point-measurement tools. You touch a probe to a terminal you suspect. Voltage, current, resistance, maybe temperature at that single point. That's excellent when you know where to look. But if you don't know where to look, the meter won't help you find it faster.
A thermal imager is not a replacement for that meter. I'd never advise anyone to leave a 116 or a 117 at home. But thermography lets you see the whole panel in one frame: the breaker running hotter than its neighbors, the lug that heats up under load, the motor that is pulling more than normal. Then you use your meter to measure specifics. They complement each other.
The mistake I see is when a technician buys a thermal camera the same way they choose a multimeter: pick the most features in a price bracket, and hope it solves every problem.
What a poor match actually costs
In March 2024, a paper mill called about a conveyor line that kept losing motors. They had replaced two 30 HP motors in one quarter and didn't want to do it again. Their maintenance electrician had been scanning the motor control center with a pocket thermal imager every couple of hours.
The problem was a power lug three bays down the cabinet. At the moment of each scan, it was not loaded enough to heat up. The line had to run under load for about twenty minutes before the connection began to show any temperature rise. A walk-around with a handheld imager, even one with excellent resolution, was never going to catch it.
What they needed was not a more detailed image. They needed a camera that would stay pointed at that cabinet and watch the temperature continuously.
The FLIR A70 thermal imaging camera is designed for that scenario. It is a fixed-mount, network-connected camera that streams temperature data and triggers an alarm when a thermal trend moves beyond a set threshold. Two days after they installed it, the failing lug showed a 30-degree temperature rise above the adjacent phases. The repair was a new lug and an hour of labor. The alternative was a third motor replacement, which would have cost thousands more plus another unplanned shutdown.
Between September 2024 and January 2025, we processed 74 rush orders. Just over a third were repeat buyers. I should add that none of their original tools were defective. The cameras were fine. The selection process was the problem. It's uncomfortable to admit from our side of the counter, because some of those first units were sold by us.
I've learned to ask what is not included before I ask what the price is. A vendor who explains the limitations of a recommendation upfront, even when it makes the sale harder, usually saves the customer more money than the discount ever does.
Choose the workflow, then choose the camera
Now when I'm triaging a rush order, I start with three questions before mentioning any specific product:
- Is this an inspection route or a single failing asset? If you're walking through a plant and checking dozens of locations, a handheld or phone-based imager is reasonable. If one asset keeps failing, a fixed camera pointed at that asset is the better return on investment.
- How far away is the smallest target? A phone-mounted sensor can be enough at arm's length, inside a panel or above a pump. The farther the target, the more spatial resolution you need to resolve it. Distance is the factor that actually pushes you toward higher-resolution sensors.
- Does the result need to be documented? If you're filing reports or proving an intermittent condition, you need images you can compare over time, with radiometric data intact. If it's a quick self-check before you leave a site, a simple image is enough.
For an HVAC tech who carries a 116 or a 117 and works on close-up equipment, such as panels, compressors, and blowers, a compact imager in the same bag can be the right tool. For a facility where the same breaker or motor keeps failing, the A70 earns its place because it does something no handheld can do: it watches when nobody is there.
These recommendations assume typical commercial and industrial conditions. If you're working in hazardous areas requiring certified equipment, or if you're measuring reflective surfaces like polished metal, the rules change. That's when a generic answer stops applying.
Back to the Friday phone call. The facilities manager didn't need the sharpest consumer thermal camera in the world. He needed to understand a pattern, why that chiller was tripping repeatedly. I recommended a camera that could stay in the chiller room and log temperature continuously while the electrician measured current with a meter he already had.
A thermal camera is a tool for finding problems. But it can only find the ones it's in front of when they happen. Resolution matters, but presence matters more. That is the question worth asking before you buy: not "how many pixels," but "will this camera be looking at the right thing at the right time?"
