About Us

Our AI-driven greenhouse, in action

Our Story

From Mae Win subdistrict to a repeatable model

Our pilot site sits in Mae Win subdistrict, Mae Wang district, Chiang Mai — a region where villages grow and process large volumes of coffee, tend kitchen gardens that supply vegetables like lettuce and tomatoes to nearby buyers, and cultivate rice across nearly every available plot.

Many villages already use semi-greenhouse structures to dry their coffee harvest. But as rainfall grows more erratic, that manual, unpowered approach is no longer enough — beans and grain are spoiling before they can dry, and too much land is being used inefficiently just to hedge against the weather.

We are currently researching and field-testing the system with two science faculty at Amnuay Silpa School before scaling it to more villages.

The System

One greenhouse. Two modes. Switched automatically.

An Arduino-based controller reads eight sensors in real time and drives six actuators — switching seamlessly between growing crops and drying harvested produce.

Full climate control for year-round vegetable and specialty crop cultivation inside the greenhouse.

Automated irrigation triggered by configurable soil moisture thresholds
Humidity regulation via ultrasonic mist system and exhaust fans
Temperature management — fans cool, ceramic heater warms cold nights
Scheduled grow-light photoperiod cycles (e.g. 16h on / 8h off) via real-time clock
AI crop recommendations generated from multi-sensor readings
Weather-API integration — pre-emptive action before forecast rain or frost

Suitable for: Leafy greens, tomatoes, chillies, herbs, mushrooms, specialty vegetables

Real Scenarios

What the AI actually decides, minute by minute

Farming Mode

Heat Spike Response

At 13:30 the sensors read 38°C against a 32°C threshold. Exhaust fans and the mist system kick in immediately — the LCD shows “TEMP HIGH — Cooling ON.” Twelve minutes later the greenhouse has settled back to 29°C.

Farming Mode

Pre-Emptive Rain Response

A 06:00 weather check forecasts heavy rain by 14:00. By 11:00 the system pre-warms the greenhouse, suspends the midday irrigation cycle, and closes ventilation to retain heat — all before the first drop falls.

Drying Mode

Coffee Bean Drying Cycle

Freshly wet-processed Arabica beans go in at 45% moisture. The system holds 45°C with fans at full speed, recalculating every 30 minutes. The LCD reads “Est. 18h remaining” — and drops to a gentle hold once 11% is reached.

Drying Mode

Paddy Rice Drying

Post-harvest rice arrives at 24% moisture. Fans cycle 20 minutes on, 20 off — mimicking traditional sun-drying turning — capped at 40°C to protect the grain. Final reading: 13.8%, safely inside storage range.

Farming Mode

Night Frost Alert

At 04:00 a forecast flags frost risk for 05:30. The heater ramps to full power, the buzzer sounds, and the LCD displays “FROST RISK — Heater at max — Check greenhouse” — hours before it would otherwise reach the crop.

Built Frugally

Roughly a third of the cost of a commercial smart greenhouse

We use readily available sensors and an Arduino Mega controller rather than proprietary hardware — keeping the full bill of materials affordable enough for a single farming household to eventually own.