UPS Nets Huge Fuel Savings With Analytics

5 Big Wishes For Big Data Deployments

Constructive dissatisfaction. That's what UPS calls its ongoing quest for process improvement that brought about ORION, an On-Road Integrated Optimization and Navigation system that will save the shipper 1.5 million gallons of fuel in its first year.

Formally announced last week, ORION entered beta testing at one distribution center in 2009. Eight more distribution centers were added in 2010 and by the end of this year it will be live and in production in 82 distribution centers serving 10,000 routes -- the daily runs traveled by UPS drivers. Getting to all 55,000 routes will take until 2017, but ORION started with rural and suburban routes on which the miles really rack up.

ORION builds on UPS Package Flow technology, which rolled out from 1997 to 2003 and used analytic modeling to determine the precise delivery order for packages along predetermined routes. Package Flow also set the loading plan for each delivery truck so drivers could quickly find each package.

What Package Flow didn't do was give drivers precise, optimized driving directions from point to point calculated to reduce driving distance, time and, thus, fuel consumption and wear and tear on vehicles. Package Flow was also fixed, meaning plans were set each morning and weren't adaptable. But UPS needed to support a MyChoice service introduced in 2011 whereby customers could take control over delivery preferences by setting delivery-time windows, rerouting deliveries to new locations, or asking UPS to hold packages for pickup at a distribution center.

[ Want more on geospatial analysis and optimization? Read Hailo Taxi App Taps NoSQL Analytics. ]

At the heart of ORION is telematics technology, which is geospatial mapping information combined with real-time communications with drivers through their DIAD (Delivery Information Acquisition Device). Drivers now get precise, optimized driving directions on their DIAD each morning, but those directions can change over the course of the day as customers request specific drop-off times, new drop-off destinations and new pickups.

The research behind ORION began in 2003, with the analytic optimization work taking years to figure out. The models had to incorporate the expected measures of time and distance, as well as business rules for service expectations; for instance, favoring high-value commercial customers and meeting MyChoice delivery windows. The model also has to be dynamic because package volumes and delivery addresses change from day to day, so routes have to adapt for maximum efficiency.

As ORION is deployed to specific distribution centers, standard mapping data is refined with up-to-date information reflecting changes to local streets and traffic patterns. ORION also incorporates precise delivery details at each address. UPS knows, for example, when deliveries are taken at the side or back of a building as opposed to the front door.

"We take all this information and we run it against our mathematical models so that every day we can provide to every driver an optimized delivery route," says Juan Perez, UPS' VP of information services. "As the drivers are completing their work throughout the day, they use the manifest displayed on their DIAD to guide them to their stops in an optimized order and route."

UPS says its proprietary ORION algorithm incorporates nearly 1,000 pages of code and evaluates more than 200,000 ways to run a single route. The bottom line, says Perez, is a three-fold payoff in better service for customers and improved operational efficiency and sustainability for UPS. For every one mile of driving that UPS can eliminate for each driver each day, the company figures it can save $50 million each year.

UPS isn't disclosing how many miles it expects ORION to save by the end of the year or by the completion of the rollout in 2017, but Perez said it will certainly surpass one mile per driver per day. As for that 1.5 million gallons of fuel saved this year alone, that translates to 14,000 metric tons of CO2 emissions eliminated -- another step toward reducing greenhouse gas emissions.

Emerging software tools now make analytics feasible -- and cost-effective -- for most companies. Also in the Brave The Big Data Wave issue of InformationWeek: Have doubts about NoSQL consistency? Meet Kyle Kingsbury's Call Me Maybe project. (Free registration required.)