Objective:
Work to develop a proof-of-concept leak detector designed to shut off a water line in the event of a leak.
Responsibilities:
Supervise the team and ensure that major project milestones were being achieved on time. Was also integral in part selection and approving final design decisions.
Tools:
Machine Learning, Sensor and Actuator Technology, Raspberry Pi
Timeline:
8 months
Project Summary

This project was done in teams of 4. We set out to address the shortcomings of existing flood detectors, which often react too late and require multiple scattered, battery-powered sensors. Our goal was to create an intelligent, compact, and user-friendly solution that fundamentally redefines how we detect and respond to water leaks. Hydroguard-X is designed to be accessible and effective, offering homeowners rapid protection against water damage. We developed a centralized system that uses a single inline water flow sensor and AI algorithms to detect abnormal water usage patterns. Once a leak is identified, Hydroguard-X wirelessly alerts the homeowner and automatically shuts off the water supply. This project also aimed to harvest energy from the flow of water through an inline generator. Although this was not realized beyond energy capture, enough was built to prove that this energy could be harnessed and potentially even deployed to onboard electronics. Key features included:

  • An inline shutoff valve capable of an immediate response in the event of a leak.
  • Flow, pressure, and temperature sensors capable of detecting abnormal behaviour.
  • AI algorithms capable of detecting leaks.
  • Alerts delivered via wireless communication.
    • Capstone Demo Photo

    Final Presentation

    Overall, this project brought together mechanical, electrical, and software disciplines. We designed a custom enclosure and implemented smart, responsive control software that was more than capable of detecting abnormalities in flow.