Leak Lock
Smart HVAC Fault Detection and Correction
Smart HVAC Fault Detection and Correction
Full Project Documentation: @github
LeakLock is a smart, affordable HVAC fault detection system designed to identify duct leaks and airflow inefficiencies in real-time. Built using ESP32 microcontrollers, distributed pressure and temperature sensors, and a web-based dashboard, it enables intelligent zone-based regulation across a three-zone mock HVAC duct. OTA updates, Wi-Fi connectivity, and live dashboard control make it scalable and user-friendly. LeakLock operates under $50 and offers an alternative to expensive commercial systems, making HVAC optimization accessible to both residential and developing regions.
LeakLock aims to solve the often-overlooked problem of HVAC inefficiency due to undetected faults. This embedded system monitors airflow and temperature data, then responds with fan control and damper adjustments to restore balance. The solution uses off-the-shelf hardware and a custom interface built with HTML/CSS/JS, focusing on simplicity and affordability compared to commercial alternatives.
HVAC systems lose energy due to hidden inefficiencies like duct leaks and sensor failures. Most current solutions are expensive, require complex setup, and lack real-time response. LeakLock addresses these limitations by offering a modular, responsive, low-cost alternative for multi-zone HVAC diagnostics.
Multi-zone temperature and pressure monitoring
Real-time sensor feedback and response
OTA firmware updates over Wi-Fi
Live dashboard for monitoring and manual overrides
Target system cost under $50
Budget limited sensor precision (e.g., low Pa sensitivity in LPS25HB)
Component integration required space optimization
I²C channel limits required a multiplexer
Environmental simulation with 3D printed ducts lacked real-world turbulence/humidity
Software written in Arduino C++ with modular FreeRTOS tasks
Sensors used I²C protocol, communication handled via WebSocket (RFC 6455)
PCB followed best practices in bypassing, grounding, and 3.3V regulation
Fail-safes included timeout-based actuator shutdowns and startup defaults
System latency: <150ms round trip
FreeRTOS improved responsiveness and multitasking
Sensor precision limited by hardware
UI does not currently support trend logging or analytics
Mechanical issues with damper integration in 3D duct
ML model failed due to lack of labeled datasets (43.6% error)
I²C misreads from misinterpreting 16-bit register format
Early complexity managing multiple sensors and task flows
LeakLock brought together hardware, firmware, UI, and real-world design principles into a system that’s scalable and cost-effective. Despite setbacks like sensor limitations and incomplete ML models, it stands as a robust foundation for future HVAC optimization tools. Future directions include cloud integration, mobile apps, better prediction models, and improved mechanical integration.
Chvala, W. D. (2021). Duct leakage in residential homes (FY1024). University of Florida IFAS Extension. https://edis.ifas.ufl.edu/publication/FY1024
Market Growth Reports (2025). Leak Detection Equipment Market Forecast. https://www.marketgrowthreports.com/market-reports/leak-detection-equipment-market-111538