Growing of mushrooms demands a close regulation of climate conditions including temperature, humidity, CO 2 level, light intensity, and moisture in the substratum. Nevertheless, a lot of farms still use manual control and reactive changes, which result in irregular production and the waste of resources. In a bid to overcome such issues, the current paper presents TwinGrow, a cloud-based AIoT Digital Twin that will help to monitor the environment of mushroom cultivation in real-time and predict as well as control its conditions automatically. The system will have a combination of IoT sensors like DHT22, MQ-135, LDR, and soil moisture sensors and ESP32 microcontroller as the edge controller. Sensor data is sent to AWSiot core, placed in dynamoDB and run on machine learning models with AWS sage maker. Digital Twin developed using AWS TwinMaker is a visualization and predictive environmental will replica of the cultivation chamber. The controlled use of mist generators, fans, and pumps, on a closed-loop system, are controlled by automated applications using AWS Lambda and Node-RED. The efficacy of AIoToriented smart agriculture using experimental outcomes indicates better environmental conditions, low water and energy usage, and predictable yields.
AIoT, Digital Twin, Smart Agriculture, Mushroom Cultivation, Cloud Computing, IoT Sensing, Predictive Analytics