Fleet telematics project (under NDA)
A real-time GPS telemetry pipeline for connected vehicles
A Node.js TCP server that ingests Teltonika tracker data, decodes it and streams positions into Redis for live maps and reports.
Outcomes
- 01Direct TCP ingestion of Teltonika device protocols
- 02Decoded positions and events available in near real time via Redis
- 03Export tooling for reporting and downstream systems
- 04Containerised deployment that is simple to operate
The challenge
GPS trackers speak low-level binary protocols over TCP. The project needed its own ingestion layer — reliable, observable and fast enough to power live maps — instead of depending on a third-party gateway.
The approach
- Protocol-first. Implemented Teltonika codec parsing, acknowledgements and connection handling on a Node.js TCP server.
- Redis as the live layer. Latest positions and event streams written to Redis so apps and dashboards read without hitting the decoder.
- Design reviews with the team. Key choices — data model, retention, export format — agreed with the people who run it.
- Ops from day one. Docker images, logging and health checks so the pipeline can be run by someone other than its author.
The result
A lean, self-owned telemetry backbone that feeds live tracking and reporting, and gives the product full control over its vehicle data.
Replace this text with your own details and real metrics (devices connected, messages per second, latency).
Next step
Got a system that's holding you back?
Book a free 30-minute call. You'll leave with a clear opinion on what to migrate, what to automate, and what to leave alone.