The technology that UPS uses to optimize its routes and achieve big savings shows how powerful data analytics could be across the public sector.
Advanced data analytics have created a heightened awareness of how marginal reductions in time or cost, seemingly insignificant at the granular level, can add up to big savings over time. Courier services know this well, and could teach the public sector some lessons on efficient fleet management.
A UPS truck driver, for instance, makes about 120 deliveries per shift. On average, the company's fleet of brown trucks makes a total of 16 million deliveries each day, which translates to an imponderable number of possible routes. To meet this challenge, a team of engineers and mathematicians in Operations Research at UPS worked for nearly a decade to build ORION -- the On-Road Integrated Optimization and Navigation program. ORION is a sophisticated algorithm that ensures that UPS vehicles take the most time- and energy-efficient routes while making multiple deliveries.
Jack Levis, the company's senior director of process management, oversees the program. He says that ORION saves each driver between seven and eight miles a day, adding up to an annual reduction of 100 million miles driven and 10 million fewer gallons of fuel used. That accounts for $300 million to $400 million in annual savings, and 100,000 metric tons less of CO2 emitted into the air each year.
What makes ORION unique is the way that it squares computational cost-cutting with qualitative measures of success, including route consistency and risk reduction. Truck drivers prefer consistent routes when they set out each day. Customers prefer consistent delivery times. "ORION is successful because we built the algorithms and tools into processes that planners, drivers and customers already use every day," Levis says. "Stability and usability are crucial. We hid the complexity from the user."
ORION is revolutionary not only because it makes routes as short as possible but also because a critical element of its logistical engine is human preferences. Government could use a dose of this functionality and customer service attitude in its logistics-heavy public services.
Waste management is a prime example. The gas mileage that garbage trucks get is in the single digits, and trash collectors make hundreds of stops each day. New York City, for instance, has more than 2,200 collection trucks and spends $2.3 billion a year collecting trash. The city spends even more to dispose of it: A nationwide network of trains, barges and trucks carries some 6 million tons of the city's garbage over 40 million miles every year. Imagine if this operation were suffused with the algorithmic power that has mapped the movement of the UPS fleet with such tremendous specificity.
Many cities already are using data to improve their garbage operations. Boston, Philadelphia and Raleigh, N.C., for instance, have high-efficiency, self-powered bins that compact trash and wirelessly alert the appropriate city agency when they are full. Enevo, a European waste collection startup now expanding into North American cities, is using sensors to measure fill levels and deliver optimized routes to sanitation drivers on a mobile tablet. Brian Pompeo, Enevo's vice president of North American sales, says that the fill data creates a previously unavailable baseline that enables an ongoing evolution of performance improvement in waste management.
These are exciting developments, but perhaps these emerging practices will soon be paired with more complex informatics, such as data on how busy a street is at a given time of day, how much a given neighborhood recycles, where the trash goes after it's collected -- the kind of auxiliary data that has helped make ORION so successful.
We are beginning to witness an enormous uptake in Internet of Things technology at the municipal level. Environmental and GIS data are helping cities tap into a well of information that can be used to make better decisions and automate more efficient practices. As these improvements grow and scale, government should ensure that these technologies not only cut costs but, more broadly, make progress in improving citizens' quality of life.
Craig Campbell, a research assistant at the Ash Center for Democratic Governance and Innovation at the Harvard Kennedy School, contributed research and writing for this column. This article was originally published on Governing.