Sensor-Integrated Automation System for Smart Indoor Mushroom Farming

Abstract

Oyster mushroom (Pleurotus spp.) fruiting is unusually intolerant of microclimate drift. A few hours outside the fruiting band is enough to distort pinning, thin the flush, and open the substrate to competing moulds. Growers working at small and medium scale still manage this drift by hand, and the resulting yield spread is largely a supervision problem rather than a biological one. This paper reports an embedded, network-connected controller that removes the operator from the regulation loop. Five sensing channels: dry-bulb temperature, relative humidity, carbon-dioxide concentration, substrate moisture and illuminance are polled by an ESP32 node, evaluated against species-specific fruiting bands, and translated into relay commands for a mist humidifier, an extraction fan, a resistive heater, and an LED grow bar. A hysteresis band is imposed on every control channel so that actuators do not chatter around the set-point, which is the failure mode most commonly reported for naive threshold logic. Over a full cropping cycle, the enclosure held temperature at 21–24 °C, relative humidity at 82–88 % RH, carbon dioxide at 850–1100 ppm and substrate moisture at 62–68 %. Measured regulation tolerance was ±1 °C, ±2 % RH and ±50 ppm; sensing latency stayed between two and three seconds and actuator engagement below five seconds, with 98 % uninterrupted operation across the observation window. Telemetry forwarded to a cloud dashboard every five seconds let the grower audit the chamber remotely and intervene on evidence rather than on inspection. The contribution is therefore not a new control theory but a demonstrably deployable one. Component cost, wiring complexity and configuration effort are all held low enough for a grower without electronics training, while the modular architecture leaves defined attachment points for machine-vision disease screening, harvest-window forecasting and photovoltaic supply.