Building a Custom AI System to Optimize Route Planning for a US Logistics Company
Logistics
4 min read
Apr 10, 2025
About the Client
Pathwise Freight is a national logistics company based in Chicago, USA, managing daily shipments across 12 states. Their operations involve coordinating a fleet of over 400 delivery trucks, with time-sensitive routes and constantly shifting traffic, weather, and demand conditions.
The Challenge
Pathwise was using a legacy route optimization system that failed to adapt to real-time data. This resulted in suboptimal routes, fuel inefficiencies, and delayed deliveries. Manual intervention was frequently needed, and the ops team had limited visibility into forecasted traffic or demand patterns.
Static routing system with no real-time adaptability
Inaccurate delivery time estimates
Wasted fuel and driver hours due to inefficient routing
No predictive insights for traffic or shipment volume
The Solution
Machine developed a custom AI-powered routing engine tailored to Pathwise’s internal systems. It integrated with GPS, weather APIs, traffic feeds, and delivery schedules. The system used real-time data and historical patterns to dynamically suggest optimal routes, flag risk areas, and forecast delivery windows.
Built a custom AI engine using Python and TensorFlow
Integrated with fleet GPS systems and traffic/weather APIs
Real-time dynamic routing adjustments
Predictive analytics dashboard for operations managers
API access for internal logistics tools
Key Results
17% reduction in fuel usage in first 60 days
24% faster average delivery times
31% decrease in manual routing changes
Improved forecasting accuracy by 40%
ROI achieved in 3.5 months
Why It Worked
Off-the-shelf solutions couldn’t meet the specific needs of Pathwise. By developing a fully custom AI system, Machine gave the company complete control, real-time intelligence, and measurable gains. The software fit into their existing workflows, but added a layer of predictive intelligence that transformed daily operations.
