> Markdown version of [/videos/1999-tracking-vehicles-at-scale](https://www.wearedevelopers.com/videos/1999-tracking-vehicles-at-scale). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Tracking vehicles at scale

Trackonix bypassed complex microservices to track fleets at massive scale. Pushing logic to PostgreSQL and leveraging client-side DuckDB-Wasm dropped their analytics server costs to absolute zero.

- **Speakers:** [Thorsten Riess](https://www.wearedevelopers.com/@thorsten-riess)
- **Event:** World Congress 2026 Europe - Virtual Stage
- **Published:** July 2, 2026
- **Duration:** 30:33
- **URL:** https://www.wearedevelopers.com/videos/1999-tracking-vehicles-at-scale

## Summary

Trackonix, a fleet management platform handling massive, heterogeneous telematics data, prioritizes architectural simplicity to preserve their small engineering team's attention. Instead of adopting complex, multi-service architectures common at hypothetical web scale, they deploy a lightweight infrastructure stack using HashiCorp tools like Nomad and Terraform over Kubernetes. At the data layer, they rely on PostgreSQL with PostGIS—hosted on Neon for serverless compute and copy-on-write branching—as a rigid single source of truth, aggressively minimizing moving parts by pushing computation directly to where the data lives.

By leveraging advanced PostgreSQL features, the team successfully eliminates entire categories of specialized databases and stream processors. They implement a Kalman filter for real-time GPS noise reduction directly in PL/pgSQL, accepting a manageable database load tradeoff to bypass stateful streaming infrastructure. Similarly, near-real-time trip detection relies on database triggers and continuous queries to merge ignition events into actionable analytics. Complex temporal graph problems, such as tracking multi-source vehicle-trailer couplings, are managed using interval columns, recursive CTEs, and deferred check constraints. This ensures that impossible states and cyclic relationships are strictly unrepresentable at the storage layer without needing a dedicated graph database.

To manage the ever-growing volume of historical telemetry, Trackonix bypasses traditional server-side data lakes by exporting aged PostgreSQL partitions to S3 as open-format Parquet files. Client browsers then query this cold data directly using DuckDB-Wasm via signed URLs, dropping server-side analytics computing costs to zero. This architecture highlights critical engineering lessons: systems should be engineered for actual scale rather than conference-slide scale, the true cost of any component is its long-term operational maintenance, and utilizing open formats prevents vendor lock-in even when relying on managed cloud services.

**Keywords:** fleet management telematics, postgresql database architecture, postgis spatial engine, hashicorp nomad deployment, serverless postgresql neon, pl/pgsql kalman filter, real-time trip detection, time-variant graph data, recursive cte queries, deferred check constraints, stateful stream processing alternative, parquet data lake, duckdb-wasm browser analytics, cold storage s3 querying, telemetry infrastructure scale

## Chapters

1. **Managing fleet telemetry complexity at production scale** (00:03) — How a small team handles heterogeneous vehicle data and peak ingestion rates without over-engineering.
1. **Adopting a single source of truth architecture** (03:23) — Minimizing integration points and reconciliation overhead by keeping logic close to the primary data store.
1. **Running operations with single-purpose infrastructure tools** (05:56) — Using discrete HashiCorp binaries for infrastructure management to avoid the complexity of full orchestration platforms.
1. **Leveraging managed serverless databases for primary storage** (07:37) — Consolidating analytical and spatial workloads into a single transactional database with compute and storage separation.
1. **Implementing stateful streaming logic inside a database** (10:31) — Replacing external stream processors with recursive SQL functions to run Kalman filters directly on incoming telemetry.
1. **Real-time trip sessionization using database triggers** (15:40) — Generating continuous session aggregations from raw events without a message broker or consumer queue.
1. **Resolving temporal graph structures with relational tables** (18:25) — Enforcing conflict resolution and cycle detection for moving hierarchical relationships using standard SQL constraints.
1. **Querying cold storage directly from the browser** (23:54) — Offloading historical data access to object storage and client-side analytical engines to bypass backend query layers.
1. **Evaluating architectural tradeoffs and final system lessons** (26:41) — Assessing the costs of running heavy computation inside the database and defining guidelines for system scaling.

## Related Moments

- [Time-series data management with PostgreSQL and TimescaleDB](https://www.wearedevelopers.com/videos/1990-flex-your-energy-building-a-cloud-native-platform-for-renewable-energy-communities) (from "Flex your Energy: Building a Cloud-Native Platform for Renewable Energy Communities")
- [Data structures, unique identifiers, and platform scaling](https://www.wearedevelopers.com/videos/473-fleet-management-reinvented) (from "Fleet Management - Reinvented")
- [Inspecting a real-time fleet management telemetry dashboard](https://www.wearedevelopers.com/videos/178-industrializing-your-data-science-capabilities) (from "Industrializing your Data Science capabilities")
- [Managing high-traffic infrastructure without chasing technology hype](https://www.wearedevelopers.com/videos/1817-how-to-avoid-tech-hype-traps-josip-stuhli) (from "How to Avoid Tech Hype Traps - Josip Stuhli")
- [Building a custom vehicle telemetry platform on AWS](https://www.wearedevelopers.com/videos/473-fleet-management-reinvented) (from "Fleet Management - Reinvented")
- [Modern application stacks and real-time data requirements](https://www.wearedevelopers.com/videos/806-leveraging-real-time-data-in-fsis) (from "Leveraging Real time data in FSIs")

## Related Articles

- [Making Data Warehouses Fast: A Developer’s Story](https://www.wearedevelopers.com/magazine/107-making-data-warehouses-fast-a-developer-s-story)
- [Why Event-Driven Architecture Isn’t About Speed (and When You Actually Need It)](https://www.wearedevelopers.com/magazine/745-why-event-driven-architecture-isn-t-about-speed-and-when-you-actually-need-it)
- [Dev Digest 139 - Soft and hard queries](https://www.wearedevelopers.com/magazine/487-dev-digest-139-soft-and-hard-queries)
- [MLops – Deploying, Maintaining And Evolving Machine Learning Models in Production](https://www.wearedevelopers.com/magazine/115-mlops-deploying-maintaining-and-evolving-machine-learning-models-in-production)

## Related Jobs

- [Lead Software Engineer - Data Engineering](https://www.wearedevelopers.com/jobs/ext/2000968-lead-software-engineer-data-engineering) at **Dynatrace**
- [Senior Software Engineer, Data](https://www.wearedevelopers.com/jobs/48273-senior-software-engineer-data) at **Sportradar Media Services GmbH**
- [Staff Software Engineer](https://www.wearedevelopers.com/jobs/ext/1425755-staff-software-engineer) at **GitHub**
- [Principal Software Engineer, Database Infrastructure](https://www.wearedevelopers.com/jobs/ext/1465908-principal-software-engineer-database-infrastructure) at **GitHub**
- [Artificial Intelligence (AI)](https://www.wearedevelopers.com/jobs/ext/1952055-artificial-intelligence-ai) at **Twilio**
- [Data Scientist](https://www.wearedevelopers.com/jobs/ext/1351648-data-scientist) at **Almedia**