AI/ML Route Optimization for a Freight Delivery Service
Challenge
The client is a US freight and parcel delivery company serving both B2B and B2C customers.
Its last-mile costs were high and routes were inefficient, with different drivers repeatedly visiting the same areas in vehicles that did not match shipment sizes. Uneven order flows made it hard to balance scheduled corporate jobs, urgent private requests, and international parcels, and dispatchers, drivers, and customers lacked a shared view of order status.
The client is a US freight and parcel delivery company serving both B2B and B2C customers.
Its last-mile costs were high and routes were inefficient, with different drivers repeatedly visiting the same areas in vehicles that did not match shipment sizes. Uneven order flows made it hard to balance scheduled corporate jobs, urgent private requests, and international parcels, and dispatchers, drivers, and customers lacked a shared view of order status.
Solution
SumatoSoft built a unified web and mobile platform that brings orders from online stores, individual requests, and international shipments into one format.
AI-driven route planning uses maps, traffic, weather, and time windows to group and sequence stops, estimate arrival times, and reroute during the day. A machine learning model recommends a vehicle class from each shipment's dimensions and weight. Dispatchers get a daily status view with delay forecasts, drivers get a mobile app with optimized stops, updated ETAs, and proof of delivery, and managers get dashboards on mileage, on-time performance, and utilization.
SumatoSoft built a unified web and mobile platform that brings orders from online stores, individual requests, and international shipments into one format.
AI-driven route planning uses maps, traffic, weather, and time windows to group and sequence stops, estimate arrival times, and reroute during the day. A machine learning model recommends a vehicle class from each shipment's dimensions and weight. Dispatchers get a daily status view with delay forecasts, drivers get a mobile app with optimized stops, updated ETAs, and proof of delivery, and managers get dashboards on mileage, on-time performance, and utilization.
Results
- 98% on-time delivery within agreed windows
- 35% decrease in fleet downtime
- 22% reduction in fuel consumption and last-mile costs
- 18% less total delivery mileage and 45% fewer repeat trips to the same area
- 40% fewer delays and 30% faster response to urgent daytime requests
- 98% on-time delivery within agreed windows
- 35% decrease in fleet downtime
- 22% reduction in fuel consumption and last-mile costs
- 18% less total delivery mileage and 45% fewer repeat trips to the same area
- 40% fewer delays and 30% faster response to urgent daytime requests