Truck-Safe at Scale: How a Global Fleet Leader Replaced Its Legacy Routing platform and transformed operations with NextBillion.ai

How a global fleet intelligence company with 4.5 million connected devices replaced its legacy distance matrix infrastructure with NextBillion.ai’s truck-safe routing — migrating 100% of production traffic and expanding into Europe while maintaining sub-second response times at enterprise scale.

AT A GLANCE

• 4.5 million+ connected devices globally
• 169,000+ nightly optimization requests processed
• Under 0.1% of requests exceeding 5-second response threshold
• 100% of production traffic migrated to NextBillion.ai

BACKGROUND

This company is one of the world’s leading fleet management and telematics platforms, serving enterprise and mid-market customers across North America, Europe, and beyond. Its platform spans ELD compliance, vehicle tracking, dispatch, route optimization, and in-cab navigation — with a customer base spanning everything from light-duty commercial fleets to heavy trucking operations with strict regulatory routing requirements.

At the core of its route optimization product is a Distance Correlation Matrix (DCM) — the engine that calculates point-to-point distances and travel times across its customers’ delivery and service networks. For over a decade, the company had generated this matrix internally. But as its customer base grew and its product ambitions expanded, the limitations of that internal infrastructure became increasingly difficult to work around.

THE CHALLENGE

The company’s internal DCM was built on OpenStreetMap data — a strong foundation for general routing, but one that lacks the truck-specific attribute layer that commercial fleet operators require. Height clearances, weight limits, axle restrictions, and hazmat routing rules are not part of OSM’s data set, meaning the company’s DCM was producing routes that were potentially non-compliant for heavy vehicles and were failing to reflect the true travel time differences between passenger and commercial vehicle routing.

This gap had become a strategic issue. Customers operating mixed fleets of light, medium, and heavy-duty vehicles needed routing that could correctly differentiate between them — and the company recognized that continuing to generate its own DCM without truck-safe data was both a product limitation and a growing engineering burden.

Key Pain Points

• Internal DCM built on OSM lacked truck-safe routing attributes: height restrictions, weight limits, axle load, and hazmat classifications
• Travel time estimates for heavy vehicles were materially inaccurate without truck-specific road network data or real-time traffic conditions, eroding customer trust in route quality
• In-cab navigation product was calling the Directions API after every GPS refresh rather than using SDK-based caching, making the product too expensive to launch at scale
• European expansion required truck-safe routing coverage in a new geography with a separate infrastructure footprint

THE NEXTBILLION.AI SOLUTION

Following an initial discovery phase, a proof-of-concept period where production traffic was replicated into a staging environment, and a phased migration across geographies, the company completed a full transition of its optimization infrastructure to NextBillion.ai — replacing its legacy internal DCM environment entirely.

Truck-Safe Distance Matrix API

• Full truck routing attribute support: vehicle height, weight, axle load, and hazmat restrictions, powered by TomTom’s commercial road network data — the same data source the company had identified as the market leader for this attribute set
• Fast API for synchronous, latency-sensitive requests
• Flexible API for batch and multi-day planning requests where latency tolerance is higher
• Global coverage across North America, Europe, and APAC markets

Navigation SDK

• SDK-based navigation replacing the previous approach of calling the Directions API on every GPS refresh — dramatically reducing API call volume and making the product commercially viable to launch at scale
• In-app integration capability: navigation embeds within the company’s existing driver application rather than requiring a separate interface
• Route deviation detection and automatic recalculation built into the SDK

RESULTS

Enterprise-grade performance at scale — Of 169,000+ requests processed in a single overnight peak window, fewer than 155 exceeded the 5-second threshold — a sub-0.1% exception rate. The vast majority of calls return in under one second.

Truck-safe routing live in Europe — Following North America, truck-safe Distance Matrix was deployed across Europe, enabling the company’s European market expansion with commercially compliant routing from day one.

Navigation on a viable path to scale — Transition to SDK-based navigation resolved the API overconsumption issue that had blocked broader rollout.

Engineering effort redirected — The engineering team redirected effort toward higher-value product work including appointment scheduling, fleet-type differentiation, and customer-facing route quality improvements.