Kalman-Powered Tracking and Geofencing

Accurate real-time location tracking is critical for applications ranging from fleet management to personal safety. Conventional GPS systems, however, suffer from errors caused by signal multipath effects, atmospheric interference, and urban canyon distortions. This paper presents a low-cost, cloud- integrated GPS tracking system augmented with a Kalman filter and dynamic geofencing designed to mitigate these limitations. Leveraging an ESP32 microcontroller and a u-blox NEO-6M module, the proposed architecture achieves position updates at 1 Hz while filtering sensor noise using a computationally efficient Kalman implementation. Field trials demonstrate a 62% reduction in median absolute positional error compared to raw GPS data, consistently achieving sub-3-meter accuracy in urban environments. Processed coordinates are transmitted to a Firebase Realtime Database via Wi-Fi, enabling live visualization on a web interface with path-history mapping, where previous coordinates are displayed as a trail on the map, and user- defined geofencing. Circular geofences are dynamically monitored using the Haversine formula, triggering real-time alerts when boundary breaches occur. The system’s responsiveness is validated through latency measurements (<500 ms end-to- end delay), with energy consumption optimized to 85 mA during active tracking modes. By integrating low-cost hardware, adaptive filtering, and cloud analytics, this work provides a scalable solution for IoT applications such as logistics surveillance and emergency response systems, addressing gaps in both precision and affordability for real-time location- based services.