The emblematic brown trucks of United Parcel Service seem to be everywhere, their stodgy appearance belying an array of high tech inside that helps UPS deliver parcels more efficiently, save gas, reduce emissions, increase safety, and minimize downtime and expenses due to repairs. Like other logistics firms, UPS continually seeks ways to optimize efficiency and expenses; it's part and parcel of the job, and it's an approach that UPS calls "constructive dissatisfaction."
But CIO Dave Barnes looks for novel ways to gain those advantages. His latest technique: adding new forms of telematics -- remote sensor-based analytics -- to the brown trucks to get new data on driver and vehicle operations that shed insight on what more can be done. "We've used telematics in other venues for years," he notes, such as to keep logs of engine condition and mileage to help schedule preventative repairs.
[ Discover what insights you can take advantage of from the other 2008 InfoWorld CTO 25 winners. ]
In 2005, Barnes initiated an R&D effort to see how UPS could increase package delivery efficiency by tracking vehicle speed, location, routes taken, idle time and so on to identify suboptimal practices. Analysis of one factor -- idling time -- helped the company reduce idling by 24 minutes per driver per day for an annual fuel savings of $188 per driver.
In one R&D effort, Barnes added more instrumentation than he knew he needed, in hopes that new data might lead to new efficiency opportunities. "This was a perfect opportunity to look at what you can do with data you didn't have before," he says.
For example, UPS investigated increasing engine measurements to better predict breakdowns so as to pull trucks in for repair before they failed -- after all, breakdowns on the road are expensive to fix and cause major delivery delays. In the R&D trial, Barnes' team added battery discharge sensors, then correlated their findings to vehicle repair histories. The effort revealed that a faster discharge rate turns out to be an excellent predictor of starter-motor failures, Barnes notes, so UPS now measures res battery discharge rates each night.
Much of that project -- which was tested out in Georgia -- is now being rolled out in stages across the company's 90,000 trucks, though R&D continues in new areas. As a result, UPS now measures all sorts of information in its trucks that it didn't use to.
UPS also began measuring driver-related activities, such as whether seat belts were engaged, related to safety. By comparing seat belt status to vehicle speed, UPS can differentiate between a driver who secures the seat belt when pulling out of a space (which is OK) and one who drives on the road unbelted (which is not OK), and then train the problematic drivers.
Similarly, UPS tracks how often drivers back up, since more accidents happen then. By combining backups (determined by wheel sensors) and location data (determined by GPS), UPS workflow analysts can then map those actions to maps and aerial photos to analyze route or parking-approach changes to minimize backups, as well as simply let backup-prone drivers know that they're increasing their risk of accidents and suggest they reconsider how they park.
"The real value isn't the information itself but what you do with it," Barnes says. That's why his team focuses as much on analytics and workflow optimization as data gathering -- and why none of this effort matters without all company groups participating and challenging each other. "There's lots of collaboration work needed across groups to realize the potential."