Mike Eskew, who served as CEO of United Parcel Service Inc. from 2002 to 2007, looked to make the company what he called a one-to-one business by treating each package as if it were the only package. His vision involved taking advantage of data and technology to streamline the delivery process, improve service and increase productivity, all while reducing costs. A major component of that technology is the handheld computers drivers began carrying in 1991 called delivery information acquisition devices (DIADs).
According to Jack Levis, UPS director of process management, this is where a suite of business intelligence (BI) tools called package flow technologies comes into play. Levis, also the program manager and business owner of package flow technologies, said the DIAD needed to become an information assistant, giving drivers better information to help them make better decisions. That was the focus of Levis’ presentation in Los Angeles earlier this month, where UPS walked away the winner of the Gartner BI Summit Excellence Award.
The problem: The rollout of the BI tools, which kicked off in 2003 and focused on the company’s U.S. small package operation, had to start at the beginning -- with the loading of vehicles. Employees who load trucks -- or pre-loaders -- memorize between 20 and 50 sheets of paper filled with codes. Those codes tell the pre-loaders what truck a package should end up on and where inside that truck it should be placed. The positioning of packages within a truck is vital, as they are shelved with ease of delivery in mind. Levis said if a driver spends an additional 30 seconds per delivery looking for a package, that can cost UPS an additional $750 million per year. Implementing a business intelligence program across the organization hinged on simplifying this process, Levis said.
The implementation: UPS needed to extract information contained in disparate corporate systems as well as external sources -- from customers in the form of electronic data exchange (EDI). UPS took all of this data and stored it in a single, integrated repository.
After consolidating the data, UPS created new labels symbolically combining internal and external information. Now, when a pre-loader packs a truck, the label is scanned, revealing information about where a package is headed and where it should be loaded onto the truck.
By digitizing the loading process, UPS could push its BI project further into the company. Before, a driver would step inside the truck, inspect the packages and determine the route for the day. Now, that package information is catalogued and contained in the DIAD. The handheld computer is also loaded with information on package delivery that used to be stored in a driver’s head. For example, it can remind drivers to wait for signatures or inform them when a customer approves leaving a package with a neighbor and another does not.
To cut down on delivery errors, UPS loaded the more than 200 million addresses it services into a database. The driver’s latitude and longitude is recorded throughout the day. That information is fed back to the DIAD, which is equipped with a GPS unit, and the handheld computer produces a visual key when a driver delivers a package to the right or wrong location. UPS can track a driver’s progress as well; every time a delivery is made, that information is shared with a central monitoring station.
Even before a package hits the pre-loading stage, dispatch supervisors use predictive analytics tools to help balance the package load of any single route a few days ahead of delivery. To do this, a package's information is sent to its delivery center before it arrives, allowing dispatchers to study the digital information, compare it to historical delivery patterns and create a forecast. A planning system analyzes that forecast to find an appropriate route. Levis said this is like “planning with a scalpel rather than an ax.” In addition, routes can be altered in midstream if a customer requests a package pickup. In some cases, Levis said, a driver’s DIAD can be updated with altered route information without human intervention.
The outcome: Levis called the implementation of analytics a “change management system.”
“For the first time in our lives,” he said, “if the data was wrong or the system fails, packages won’t get delivered.”
But the cost benefits are readily apparent. Levis reported training time for new employees has been reduced from three weeks to three hours. Because of efficiencies created by the new system, drivers are spending less time on the road (30 million miles fewer per year), gas usage is down by 3 million gallons per year, and the DIAD means manual entries have dropped by 8 billion per year.
The company also needed to get its hands on even more data. When project flow technologies was rolled out, UPS had 3% EDI with customers. Today, that figure is at 98%.
Now, Levis said, he’s rolling out a new generation of package flow technologies. This includes a product called Orion that will generate the most efficient delivery route for a driver. Currently, a driver’s DIAD contains information on all packages that need to be delivered and places them in a general order of delivery, but drivers ultimately determine the best routes possible. The new system will take away the guesswork.
Levis has, in other words, come full circle, working to digitize information from employees and looking to cut back on time spent on the road. The new generation of tools has been rolled out in a handful of locations, and Levis said the results are apparent and even longtime drivers recognize the efficiencies it’s creating.
Dig deeper on Business intelligence strategy