Farmers and geologists in developing regions cannot afford expensive spectral sensors ($300–$1,000+) needed to measure soil health or identify minerals. Existing low‑cost RGB sensors miss the infrared signatures essential for detecting iron, copper, and organic matter, leading to poor decisions and wasted resources.

The device combines a standard RGB color sensor (TCS3200) with two infrared LEDs (850 nm and 940 nm) and a temperature sensor on a single 6x8 cm PCB. An ATmega328P microcontroller controls LED sequencing and measures three data types:

• RGB reflectance under white light,
• Infrared reflectance at 850 nm and 940 nm,
• Signal Settling Time – the time (in milliseconds) the photometric signal takes to stabilize after the LED turns on.

A hybrid k‑NN algorithm fuses these nine features to classify soil types or mineral ores. Results are shown on an onboard OLED or sent via serial.

What makes it novel:

Existing low‑cost RGB sensors miss near‑infrared signatures, causing >30% errors in mineral identification. High‑end spectrometers cost 300–1000+. Our key innovation is Signal Settling Time – a new physical parameter that correlates with material density and metallic content. No other low‑cost sensor uses this temporal feature. Combined with IR LEDs, the system achieves 95% classification accuracy for iron/copper ores and soil nutrients at a production cost under $30 – a 90‑97% cost reduction.

How it would be produced:

The design uses standard, off‑the‑shelf components (TCS3200, ATmega328P, LEDs, passives). PCB is two‑layer, 6x8 cm. Assembly can be done manually for small batches or pick‑and‑place for volume. A bill of materials totals $28.70 per unit at 100+ quantity. All files are open‑source (CERN OHL v2) to enable local manufacturing anywhere.

Where it would be applied:

Precision agriculture: Smallholder farmers test soil nitrogen, phosphorus, potassium on‑site, reducing fertilizer overuse by 20‑30% and lowering emissions.

Geology / mining: Artisanal miners identify iron, copper, and other ores without lab fees.

Environmental monitoring: With added wireless (LoRa/Wi‑Fi), it becomes an IoT node for soil and water quality.

Validation:

The project has been recognised by OSHOP (Open Source Hardware Innovation Platform, Dresden) with a DOI‑archived conference paper, validated in collaboration with Fraunhofer IWU, and nominated for the German AMA Innovationspreis 2026.