Smart Plant Watering Using TinyML: Water Savings through Predictive Control
Indoor plant watering is not always effective - people often overwater or underwater plants, wasting water and harming plant health. In view of this, a smart watering system using artificial intelligence that runs on a tiny microcontroller chip has been developed. The proposed system predicts when plants need water and waters them automatically. Testing on 12 plants for 3 months has showed 27% water savings versus manual watering and 15% savings versus simple automated systems. The AI model is only 8.7 KB and runs for months on battery power without Internet.