Automatic Filter Health Monitoring in an HVAC System using FPGA based Hardware Accelerator

Occupants of enclosed residential, commercial, and industrial zones face significant health risks due to factors like temperature, humidity, and dust. In order to provide comfortable living and working environments, humidity ventilation air conditioning (HVAC) systems keep these indoor space elements stable. The air filter reduces the particulate, which improves breathability. Additionally, the filter's condition affects the HVAC system's ability to maintain humidity and temperature. In this paper, we present a novel feature-based machine learning (ML)-based hardware Field Programmable Gate Array (FPGA) implementation of an air filter health detection system. This implementation examines the sensor's real-time data to determine the state of the air filter. The detection system achieved an impressive accuracy of 96.7% while consuming only 161 mW of dynamic power.