Fleet Route Optimization and ML ETA Prediction
Supply Chain & Logistics — More than 400 deliveries a day were planned manually by dispatchers; trucks often left half full and customers received…
Client: Regional Logistics Provider. Sector: Supply Chain & Logistics.
The challenge: More than 400 deliveries a day were planned manually by dispatchers; trucks often left half full and customers received unreliable time slots.
What we built: Paired a vehicle-routing optimiser with machine-learning ETA predictions built on past GPS data and traffic patterns, delivered through a dispatcher dashboard and WhatsApp updates to customers.
Tech used: OR-Tools, XGBoost, Kafka, PostgreSQL, WhatsApp API.
- Fuel Spend: -14%
- Deliveries on Time: +19%
- Time Spent Planning: -80%
- Vehicles Needed: -11%