SwimTrack: Drowning Detection using RFIDs

This project is a blend of IOT and Machine Learning. IOT, since it invloves using RFID sensors to collect raw data and Machine Learning is used to process that data. The project aims to detect a potentially drowning person in a swimming pool using passive RFID tags. The tags are attached to the wrist of swimmer while an RFID reader placed outside the pool records readings from the tags at a high frequency. Based on the RSSI (Received Signal Strength) and frequency of detection of the tags, we model the problem as a classifier that can differentiate between swimming and drowning motion. For data collection the experiment was performed at the University of Waterloo campus swimming pool and the test subject performed multiple strokes including drowning motion.