Biometric Attendance System

A tamper-proof fingerprint attendance system — from sensor hardware and RTOS firmware to a role-based web dashboard — deployed across schools and offices with full hardware-to-dashboard delivery.

127 Fingerprint templates per device

Sub-1s On-sensor match response time

3+ Institutions deployed

Problem

Manual and card-based attendance systems are fundamentally insecure — proxy sign-ins are trivially easy and attendance records are inaccurate. The goal was a biometric attendance system deployable in schools and small offices without enterprise hardware budgets or per-device subscription licensing.

The system had to work offline (the fingerprint reader must always respond even without network connectivity), sync records reliably when the network returned, and produce per-department reports accessible to administrators without exposing all data to all staff.

Approach

Chose the R305 optical fingerprint module for on-sensor template matching — the sensor itself identifies the finger without sending raw biometric data to the microcontroller; this keeps match response under one second and means no biometric data is transmitted over the network.
Designed a dual-task RTOS firmware: one task handles fingerprint read/match loops continuously; a second task manages WiFi sync to the backend; RTOS priority ensures the reader task is never blocked by network operations.
Built 127-template storage directly in the sensor’s flash memory (the module’s maximum capacity), with a local fallback log in the NodeMCU EEPROM for records captured offline.
Developed a PHP/MySQL backend with per-department role-based filtering — department heads see only their team’s records; administrators see all; export to CSV is available per date range.
Designed and 3D-printed custom ABS enclosures in both desk and wall-mount configurations so the system could be installed without retrofitting existing infrastructure.

Outcome

Deployed across schools and office environments, the system eliminated proxy sign-ins entirely — attendance records became verifiable and consistent. Hardware cost per unit was kept below what commercial fingerprint terminals cost without any of the cloud dependency or subscription requirements. The project delivered the full stack from physical hardware to browser-based reporting.

Architecture

System design shown as an operational topology, not a decorative diagram.

The architecture visual is intended to show how edge capture, buffering, transport, event processing, and operator-facing systems fit together under real deployment constraints.

Architecture diagram showing NodeMCU dual-task layout with on-sensor fingerprint matching, offline-first logging, WiFi sync to MySQL backend, and web dashboard with per-department role-based access. — Biometric Attendance System

What to look for

Offline resilience and deterministic recovery are treated as first-class design concerns.
Operational visibility is separated from ingestion reliability so reporting never destabilizes capture.
Protocol choices reflect topology and failure modes, not just team familiarity.

127 Fingerprint templates per device

Sub-1s On-sensor match response time

3+ Institutions deployed

What I Would Do Differently

Replace the synchronous Wi-Fi sync with a local-first queue and background flush. The original approach blocked the fingerprint reader thread during sync operations — causing visible latency spikes when the network reconnected after being offline. A queued approach would keep the sensor always responsive.

Stack

The implementation choices below reflect the boundary between field reliability, event architecture, and operator-facing systems.

NodeMCU C++ FreeRTOS PHP MySQL 3D Printing