AI/MLPythonReactOR-Tools

Route Optimization System

Built a route optimization engine for a 150-vehicle fleet. 25% reduction in fuel costs, 94% dispatcher adoption.

25%
Fuel Cost Reduction
1000+/day
Routes Optimized
94%
Dispatcher Adoption

The Challenge

Route planning was done manually by a team of dispatchers using a combination of Google Maps and gut instinct. It worked, but it didn't scale — and fuel costs were eating into margins during the 2022–2023 price spikes. They tried two off-the-shelf routing tools before reaching out to us; both failed to integrate with their existing dispatch system.

What We Did

Built a route optimization engine using Google OR-Tools, integrated with their existing dispatch system via API. The key insight: dispatchers don't trust systems they can't control. We spent a month on UX for the dispatcher panel alone — letting them override or adjust routes without touching the algorithm directly. The system suggests routes, but humans have the final say.

The Result

25% reduction in fuel costs over first 6 months. 1000+ routes optimized daily. 94% dispatcher adoption rate — way higher than the off-the-shelf tools they tried before. Roughly 15% reduction in CO2 per delivery km as a side effect.

What We'd Do Differently

Integration with their legacy dispatch system took longer than expected — their API documentation was outdated, and we had to reverse-engineer some endpoints. Also, the first version of the algorithm over-optimized for fuel efficiency and created routes that were hard for drivers to follow. Added a 'route simplicity' parameter after driver feedback.

Ready for similar results?

Let's discuss how we can help you achieve your goals.

Start your project