UPS's ORION system calculates 30,000 route optimizations per minute, saving 10 million gallons of fuel annually

Background

UPS operates one of the world's most complex delivery networks, with hundreds of thousands of drivers making millions of stops daily across 220+ countries. Small inefficiencies in individual routes compound across that scale into massive fuel, time, and cost waste. Traditional route planning could not handle the combinatorial complexity of optimizing routes across the full fleet in real time. ORION was built to solve this at scale.

What Was Implemented

• ORION: AI-driven route optimization analyzing daily package deliveries, pickup times, traffic, parking, and historical route data
• 200,000+ routing options evaluated per driver daily
• 30,000 route calculations per minute at the system level
• Left-turn penalization logic to reduce idle time and collision risk
• Dynamic routing add-on (reported separately in 2022) adding real-time adaptability to weather and traffic changes, saving an additional 2-4 miles per driver

Results

ORION saves 10 million gallons of fuel annually , eliminating 100,000 metric tons of CO₂ per year. The system generates more than $300 million in annual savings against a development investment of approximately $250 million. It processes 30,000 route calculations per minute and evaluates more than 200,000 routing options per driver daily.

Lessons

• Domain-specific heuristics (e.g., penalizing left turns) can deliver large aggregate gains when applied at fleet scale — even if they appear counterintuitive at the individual route level
• Investment in AI route optimization at this scale requires patient, multi-year development; UPS invested $250 million before realizing returns
• Adding dynamic (real-time) routing on top of a static optimization foundation compounds efficiency gains
• INFORMS recognition and BSR documentation provide independent validation of vendor-reported figures