Seeing the Invisible: How we put a laboratory in a pocket.

Imagine walking down a line of growing lettuces. To the naked eye, they all look perfect. Green, crisp, healthy. But one of them is dying. You just can’t see it yet. Now, imagine if you could spot that disease four days before a human eye ever could. The impact? Massive. Farmers stop wasting water and nutrients on a lost cause. The disease doesn’t spread to the healthy crop. Yields go up. Waste goes down.

This isn’t science fiction. It’s hyperspectral imaging, capturing the full light spectrum of a pixel to identify a material’s unique chemical fingerprint. PyrOptik, a pioneer in high-temperature industrial measurement, had built a groundbreaking algorithm to do exactly this. But they had a challenge. A big one.

PyrOptik had the maths. They wanted to create ChromaMapper, a product that would commoditise this technology, making it small, affordable, and usable on a standard mobile phone. But taking an algorithm designed for high-end industrial gear and making it work on a consumer smartphone? That is a minefield of unknowns. Could a device built for Instagram selfies really survive an industrial environment?

We needed total control over the hardware. Exposure time. Autofocus. White balance. But demanding that level of performance drains power. Would the battery even survive the workload? Then there’s the data avalanche. To get accurate answers, we needed a torrent of high-resolution photos captured in rapid succession.