DIY Thermocam: From Classroom Idea to Global Product

The inspiration for my low‑cost thermal imager began in a 2010 physics class when our instructor introduced a single‑point infrared thermometer—commonly known as a pyrometer—into the curriculum. My friend and I seized the opportunity to transform that lone sensor into a full‑field thermal scanner using servo motors to sweep the detector across a wide area.

Our initial proof‑of‑concept prototype combined Lego Mindstorms with a custom data interface, running on a PC and leveraging an automated mouse‑keyboard script in Adobe Photoshop to stitch low‑resolution thermal images. The following year I refined the design into what I called “Cheap‑Thermocam V1” (Figure A). It used an Arduino microcontroller, two servos, and Java‑based software, all for a total material cost of roughly $100.

Figure A. Photo by Max Ritter

First Kits

In 2011, I received a special prize and released the hardware schematics and source code online. The response was overwhelmingly positive; numerous hobbyists built their own variants (Figure B), and several expressed interest in purchasing a ready‑made unit. Consequently, I developed the second iteration (Figure C) in mid‑2013 and began selling it worldwide. At 20 years old, this represented a significant milestone. The Cheap‑Thermocam V2 featured a small LCD, rotary‑encoder controls, and optional SD‑card storage.

Figure C

A year later I completed the Cheap‑Thermocam V3 (Figure D), which incorporated a large touch screen, a sleeker chassis, and a faster microcontroller. Versions 1–3 all relied on the same scanning principle—moving a single‑point IR sensor across the field—which, while inexpensive, required several minutes to capture a full thermal frame. For applications involving motion, this latency was unacceptable, prompting me to explore alternative architectures.

Figure D

A New Sensor

Figure E

In 2014, FLIR released their Lepton sensor (Figure E), the first low‑cost thermal array on the market. I integrated it into the next version, the DIY‑Thermocam V1. This represented a leap from the scanning approach to high‑resolution, real‑time imagery and positioned the device as a credible alternative to commercial offerings from FLIR, FLUKE, and others.

Version 2 of the DIY‑Thermocam introduces the Lepton 3.0 array, offering four‑times the resolution of its predecessor, a more powerful microprocessor, a high‑speed visual camera, and removable storage. A new video‑output module enables live streaming of the thermal data directly from the unit.

One of the key strengths of the Thermocam is its fully open‑source ecosystem. Users can adapt the firmware and PCB designs to suit custom requirements or use the platform as a foundation for their own thermal imaging projects. The on‑board firmware features an intuitive touch‑menu interface, multiple color palettes, real‑time analysis, and flexible saving options. For PC‑side processing, raw data files are compatible with ThermoVision, a powerful viewer developed by a German programmer in Berlin. A complementary Python toolkit also allows real‑time streaming and basic analysis on a desktop.

Lessons Learned

Throughout the development cycle, I encountered numerous obstacles and adopted creative workarounds. For instance, I initially considered 3D printing the enclosure, but even high‑end printers demanded extensive post‑processing, which was impractical for small‑batch production. Injection molding was cost‑prohibitive for low volumes. Ultimately, I chose laser‑cut acrylic panels—obtained from a German supplier—that required no post‑processing and delivered a clean, durable housing.

I learned that an overly aggressive timeline inevitably invites unforeseen setbacks. A pragmatic rule I follow is to double the projected schedule; finishing early is rarely a problem because there is always room for refinement.

All the knowledge required for this project came from online resources and books; there was virtually no external mentorship. Although the learning curve was steep, the experience was invaluable. Building a prototype, selecting components, ensuring ease of assembly, performing quality checks, managing global shipping, marketing, and customer support—every step demanded experimentation. The mistakes I made over the years ultimately forged the expertise I possess today. I believe that hands‑on problem solving accelerates both personal and technical growth more than formal study.