RezQ - A Drowning Detection System with Real Time Alerting and Location Tracking

Abstract

Drowning is a global health crisis and is a leading cause of unintentional injury death worldwide. A solution is urgently needed to save people's lives at the pool and beach. Currently, there is no low-cost, efficient method to optimize rescue operations for drowning. 

RezQ is a wristband that detects drowning based on heart rate and depth and automatically contacts rescue services, bystanders, and the user’s family. It provides rescue services with live information on the user’s location, vital signs, and pre-existing medical conditions. This project aims to evaluate RezQ's performance and determine whether it can be implemented on a large scale for high-risk swimmers.

The effectiveness of this device will be measured through a test to see if the device is providing all the necessary information to the rescue services and contacting the correct people and if the cost is reasonably low. The accuracy will be tested using an ML model that simulates 10,000  random depth and heart rate graphs, and RezQ will identify data as drowning or not drowning.

Using five different ML Models (KNN, CNN, XGB, Decision Tree, and Random Forest), RezQ had the best results using the XGB model with R^2 values around 85%. RezQ costs $30, accessible to roughly 78% of Americans. RezQ provided the necessary information to the rescue services, bystanders, and the user’s family. This proves that RezQ is optimal for preventing drowning in beaches and pools.