Real-Time Fleet Tracking Platform for UAE Logistics

The Challenge

A Dubai-based last-mile delivery company operating across the UAE had grown from 80 to 400 vehicles in 18 months. Their fleet management was a patchwork of WhatsApp groups, manual Excel dispatch sheets, and a basic GPS vendor dashboard that showed vehicle positions with a 5-minute delay.

No real-time visibility. Dispatch managers could not see live vehicle positions. When customers called asking "Where is my delivery?", the team had to phone drivers individually. During peak hours this meant 200+ daily calls just for status updates.

Inefficient routing. Drivers were assigned routes manually each morning. With no dynamic re-routing, a driver could pass within 500 metres of a new pickup while heading to a location 15 km away. Fuel costs had ballooned to 28% of operating expenses.

Driver accountability gaps. Without digital proof-of-delivery or automated time tracking, disputes over missed deliveries and idle time were resolved through guesswork. The company estimated it was losing 12% of revenue to unverified delivery failures and fuel misuse.

The Approach

RG INSYS proposed a three-part system: a web-based operations dashboard for dispatchers, a React Native mobile app for drivers, and an AI route optimisation engine that recalculated assignments in real time.

Live tracking infrastructure: Each vehicle's OBD-II device streamed GPS coordinates every 5 seconds via MQTT to our backend. We built a time-series ingestion pipeline using TimescaleDB on top of PostgreSQL, capable of handling 80 events per second across all vehicles without degradation.

AI-powered route optimisation: We implemented a constraint-based optimisation engine using Google OR-Tools, factoring in delivery windows, vehicle capacity, traffic patterns from Google Maps Platform, and driver shift limits. Routes were recalculated dynamically when new orders arrived or when a driver reported a delay.

Driver mobile app: Built with React Native for iOS and Android. Drivers received turn-by-turn navigation, scanned package barcodes, captured photo proof-of-delivery with GPS-stamped timestamps, and logged breaks. Offline-first architecture ensured the app worked in areas with poor cellular coverage.

Timeline: Week by Week

Weeks 1–2: Discovery with operations team. GPS data pipeline architecture. Database schema design (TimescaleDB for telemetry, PostgreSQL for business data). CI/CD and infrastructure provisioning on AWS.

Weeks 3–5: Live tracking dashboard with map view (Mapbox GL). Vehicle status indicators, geofence alerts, and historical playback. Dispatch management UI with drag-and-drop route assignment.

Weeks 6–8: React Native driver app: authentication, route view, barcode scanning, photo proof-of-delivery, offline sync. Push notifications for new assignments and route changes.

Weeks 9–10: Route optimisation engine integration. Dynamic re-routing based on real-time traffic and new order injection. Analytics dashboards: fuel consumption, delivery SLA compliance, driver performance.

Weeks 11–12: Load testing with 400 simulated vehicles. Pilot rollout with 50 vehicles. Bug fixes, driver training, and full fleet deployment.

Tech Stack

• Backend: Node.js 20, Express, TypeScript, Prisma ORM
• Mobile: React Native (iOS + Android), Expo
• Frontend: React 18, TypeScript, Mapbox GL JS, Tailwind CSS
• Database: PostgreSQL 16, TimescaleDB (telemetry), Redis 7
• Real-time: MQTT (vehicle telemetry), Socket.io (dashboard updates)
• Route Optimisation: Google OR-Tools, Google Maps Platform
• Infrastructure: AWS (ECS, RDS, IoT Core, S3, CloudFront)
• AI tooling: Claude Code, Cursor IDE

Results

Key Features Delivered

• Live operations dashboard: Real-time map with vehicle positions, colour-coded delivery statuses, geofence breach alerts, and historical route playback for any vehicle over the past 90 days.
• Dynamic route optimisation: AI engine recalculates optimal routes when new orders arrive mid-shift, factoring in traffic, capacity, and delivery windows. Dispatchers can override with manual drag-and-drop.
• Driver mobile app: Turn-by-turn navigation, barcode scanning, photo proof-of-delivery with GPS stamp, break logging, and offline-first sync for areas with weak connectivity.
• Analytics & reporting: Fuel consumption trends, delivery SLA adherence, driver scorecards, and idle time reports. Automated weekly PDF reports emailed to operations managers.