Logistics
AI-powered route optimization for Orbital Logistics
Built a machine-learning pipeline that cut fuel costs by 18% across a 2,400-vehicle fleet.
The challenge
Orbital was using legacy routing software that could not factor in real-time traffic, weather, or driver break windows. Fuel costs had grown 24% year-over-year.
Our solution
We designed an ML model trained on four years of historic routing and delivery data, integrated it into Orbital's dispatch system, and layered in real-time weather and traffic feeds. Dispatchers see recommended routes with confidence scores.
Tech stack
PythonPyTorchPostgreSQLKafkaAWS SageMaker
Results
18%
Fuel cost reduction
94%
On-time deliveries
2.4k
Vehicles in fleet
6mo
Time to payback
Next case study