The encyclopedia · Software & IT · Technical decision · 2003–2017

UPS built ORION routing to cut 100 million miles a year and won the 2016 Edelman Award.

For 55,000 drivers, ORION's metaheuristics pick the day's pickup-and-delivery route, saving UPS an estimated $300–400 million a year.

UPS

the move

In 2003 UPS began modernizing pickup and delivery with Package Flow data systems; the effort grew into ORION, which every day builds an optimized route for each of 55,000 US drivers from that day's packages.

ORION solves the routing problem with a metaheuristic and keeps routes consistent from day to day so drivers can build familiarity. Building and deploying it cost more than $295 million, and INFORMS awarded it the 2016 Franz Edelman Prize.

At full deployment UPS projected savings of 100 million miles and 10 million gallons of fuel a year, $300–400 million annually, and 100,000 tons less CO2 — roughly the emissions of 21,000 cars.

why it works

A fixed mathematical target for every route replaces driver guesswork with a searchable optimum.
Consistency constraints keep tomorrow's route familiar, so drivers accept the machine's plan.
Package Flow data gave ORION accurate, real-time inputs to optimize against.
UPS paired the rollout with change management across 55,000 drivers and executives.

the payoffOptimize every route, then change management to accept itinspired

what transfers

A huge optimization is worthless if the people who must follow it do not trust it, so pair the algorithm with the change management that makes it stick.

what came after

ORION became the foundation for newer UPS planning systems and a template for large-scale vehicle routing; in 2020 INFORMS honored UPS for longstanding contributions to OR practice. The routing-telematics platform expanded internationally, and its savings figures became standard citations in logistics research.

references

spotted an error? The archive wants to know.

same kind of clever

1996–1998 · Software & IT