PROJECT 003Electrical EngineeringCompleted
Face Recognition + Knocking Pattern Unlocking System
A smart security system that combines face recognition and a unique knocking pattern to provide secure, two-factor door access.
Engineering Specifications
Rated Power3500 VA (7000 VA peak)
Output Voltage230V AC ± 1% Pure Sine
THD< 2.1% at full resistive load
Transfer Time< 8 milliseconds
Lead:Fikru Yenialem
Partner:Ethiopia Giftedness and Talented Students Institution
Timeline:2026-08-12 — 2026-09-13
Technologies & Tools
ESP32KiCadProteusC++openCV
Current Status Overview
Deployed in daily production at the Zuhura Robotics Workshop. Web telemetry dashboard operational 24/7.
The Problem Statement
Traditional door locks rely on keys or passwords that can be lost, copied, or shared. We need a smarter way to secure door access.
Our Zuhura Engineering Solution
We developed a smart door lock that combines facial recognition and a unique knocking pattern to authenticate users before unlocking the door.
Design & Engineering Details
The system uses Python, OpenCV, and machine learning for face recognition, while Arduino and a piezoelectric sensor detect knocking patterns and control the servo motor.
Photo & Media Documentation
Solar array wiring and MPPT charge controller integration
Sine wave output analysis on digital storage scope
Testing & Empirical Results
We tested different faces and knocking patterns. The system successfully unlocks the door only when both authentication steps are correct.
Challenges & Failures Encountered
In true engineering, failure analysis is our greatest teacher.
We faced challenges with facial recognition accuracy, sensor sensitivity, and distinguishing valid knocks from accidental ones.
Future Roadmaps & Scaling
We plan to improve recognition accuracy, enhance security, and add mobile monitoring and access logs.
Engineers on this Project
Want to contribute to this project pod?
Apply to Join Team