AI-Powered Water Spray Mechanisms for Feline Repellence in Ecological Conservation

Abstract

The AI-powered feline repellent system using AlphaVision addresses ecological and public health challenges caused by domestic and stray cats. Cats are responsible for the annual loss of millions of birds and disrupting ecosystems, furthermore their faeces can transmit toxoplasmosis, posing risks to unborn babies and immunocompromised individuals. AlphaVision, which is the prototype system developed in this study, integrates Jetson Nano, Raspberry Pi Zero 2 W and MobileNet-SSD v2 model with a water-spray mechanism to deter cats humanely. The system uses real-time AI detection via a trained MobileNet-SSD v2 model to identify cats and activating a water spray upon detection. Field tests showed 82% detection accuracy under optimal conditions (bright lighting, 2-meter range) but reduced accuracy at greater distances, wider angles and low-light conditions. The project emphasizes sustainability by employing energy-efficient hardware and a power-efficient water-spray mechanism. AlphaVision demonstrates the potential of AI-based solutions for wildlife conservation and public health. Future improvements include expanding the dataset, optimizing the model and enhancing hardware to improve performance across diverse real-world scenarios.