Water Leak Detection Using Piezoelectric Sensors 

Designed and implemented a simple, non-invasive system to detect water leaks by measuring vibration signals in water pipes. The project focused on collecting and analyzing sensor data to distinguish between normal water flow and leak conditions and was experimentally tested on both metal and plastic pipes. 

Objectives

• Design an external, low-cost sensing solution for detecting water leaks without cutting or modifying pipes.

• Capture vibration signals generated by water flow and leakage using piezoelectric sensors.

• Analyze vibration signals to identify leakage signatures.

• Build a reliable experimental dataset for future AI-based analysis.

Outcomes & Contributions

• Reached the final stage of Miyahthon 2 after competing with over 200 teams.

• Observed significantly higher signal energy from sensors placed closer to the leak location.

• Identified clear differences between operating conditions:

  • No-leak: low and stable vibration levels

  • Leak: higher vibration amplitude and wider frequency content

• Validated results on different pipe materials (metal and plastic).

• Used RMS analysis to quantify vibration strength and confirm leak presence.

Technical Details & Skills

• Sensors: Piezoelectric vibration sensors

• Hardware: Arduino Uno for real-time data acquisition

• Programming: Arduino IDE (C/C++)

• Signal Processing: Time-domain analysis, FFT, RMS

• Tools: MATLAB for data analysis and visualization

• Field: Embedded Systems, Signal Processing