Agrotrack

Agrotrack is an IoT-powered geofencing platform designed to reduce conflict between farmers and livestock herders. It integrates physical tracking hardware with a digital mapping system to protect agricultural zoning:

• Virtual Farmland Geofences: Custom virtual boundaries created by farmers to monitor perimeter safety around crop zones.
• Live Telemetry Stream: Real-time map displaying GPS coordinates of grazing herds next to protected farm zones.
• Instant Perimeter Alerts: Automated SMS/push notification triggers dispatched to farmers and herders the moment livestock cross a geofence.
• Hardware Node Sync: Low-power ESP32 microcontrollers with GPS modules sending coordinates to the cloud via cellular nodes.

I engineered the mapping visualization and real-time data sync pipeline. Key contributions included:

• Interactive Geospatial Layer: Built a responsive, vector-based rendering layer with Leaflet and React to project complex geofences and path histories without lagging the DOM.
• Telemetry Sync Loop: Connected real-time database updates from IoT hardware (ESP32/SIM800L GPS) to the frontend, utilizing debounce algorithms to prevent map stutter.
• Alert Event Triggers: Set up serverless function endpoints to process incoming location payloads, calculate intersections, and queue immediate SMS alerts.

• High-Frequency Rendering Lag: Drawing dozens of active GPS tracks and multi-point geofence boundaries caused major browser reflow bottlenecks. I optimized this by implementing canvas-based map rendering rather than SVG DOM elements, cutting rendering time by 75%.
• Latent GPS Telemetry: Fluctuating cellular signals in remote areas sent duplicate or delayed GPS packages. I resolved this by coding a temporal-distance filter on the backend to verify coordinate velocity before updating the live map.