Delivery Route Optimisation

We built a routing engine to optimise daily vehicle routes, reduce travel time, and boost operational efficiency

Machine Learning

Logistics

Finding the shortest and most efficient routes is a core challenge in operational research. To optimize Sociedade de Cerveja field operations, we built an end-to-end routing solution that calculates travel durations, manages constraints and generates optimized daily routes for multiple workers.

Implementation time:

4 months

in fuel savings

€100k

locations visited per day

+10%

coverage

+35%

The Opportunity

The beverage company needed a way to optimize its routes, because they were facing

High time waste

Workers spent several hours travelling inefficiently between locations, reducing daily coverage of points of sale.

‍

Complex constraints

Routes had to respect limits such as maximum time per route, vehicle constraints and different frequency of visits

‍

No optimisation engine

There was no system capable of generating fast, cost-efficient routes or adapting to changes

‍

The Solution

We built a routing engine designed to optimize daily field operations. Using an OSRM-based server, we automatically calculate travel durations between every pair of locations and generate a complete duration matrix. With this foundation, our optimisation algorithms create routes that minimise travel time while respecting real constraints such as maximum route length, vehicle limits and repeated visits. The system then produces clear, ready-to-use schedules for one or multiple workers.

The approach is flexible: it adapts to real-time changes, tests multiple routing strategies, and selects the most efficient one. By combining classic machine learning with optimisation methods, we help operations teams reduce wasted kilometres, expand coverage, and increase predictability in the field.

‍

Meet the team of this project

Ricardo Neves

Data Scientist

Gustavo Fonseca

Tech Lead

The Impact

Since its introduction in early 2025, the routing engine has changed the way field teams plan their day. Instead of relying on manual planning or habit-based routes, teams now receive optimised schedules that help them cover more ground with less effort. With shorter travel times and smarter ordering of visits, workers can reach more locations per shift, respond faster, and operate with greater predictability.

‍

This efficiency also translates into tangible savings. Fewer wasted kilometres mean lower fuel consumption and reduced operational costs, while improved coverage ensures that teams maximise the value of every hour on the road. What began as a technical optimisation quickly became a strategic advantage, giving the client a faster, clearer, and more scalable way to manage daily operations.

‍