How U.S. Xpress uses data mining for maintenance

user-gravatar Headshot
Updated May 30, 2017

Jeff Seibenhener, chief information officer of U.S. Xpress

Every desktop PC in the IT department at U.S. Xpress has the same wallpaper image of a high-tech truck with this slogan:

Technology and Business
Together we Deliver

“It is critical that you deliver these relationships and that you build strong relationships,” said Jeff Seibenhener, chief information officer of U.S. Xpress, during a breakout session at the CCJ Spring Symposium on Tuesday, May 23, in Ashville, N.C.

During the breakout, Seibenhener and Gerry Mead, senior vice president of fleet maintenance, discussed a project they worked on to apply business intelligence and data mining to create “one stop shop” reporting for U.S. Xpress’ fleet maintenance.

U.S. Xpress used the Microsoft BI stack to convert raw data from multiple systems to intelligent information that is delivered to users, in real time, through a web-based SharePoint interface. Users at all levels can “quickly understand if we are winning or losing,” Seibenhener said.

“We are driving that business efficiency and managing both equipment and people really fast,” Mead explained.

During the breakout, Mead showed a number of reports that he and others in the company use to decrease costs and increase uptime and asset availability, thereby giving operations the opportunity to generate more revenue and eliminate disruptions for drivers.

A Tractor Velocity Report, for example, has one section “Unseated Trucks—Shop” that shows the number of trucks in U.S. Xpress, by division, that have been in the shop for more than 24 hours. Users can drill down to find out why the trucks are in the shop and take actions to improve efficiency.

Another report, Standard Repair Times, shows how efficiently the shops are completing repairs. Each type of repair is represented by a stoplight icon that is colored green, yellow or red depending on how the average repair time is trending in relation to the fleet’s SRT.

For Mead, one of the most important reports shows failures that occur between preventive maintenance events. The company’s trade cycles for tractors are 500,000 miles, and it has a PM service every 50,000 miles to change oil, fuel filters and more.

The company is 94 percent efficient, meaning less that only 6 percent of its trucks are returning to shops between PMs. Ninety percent of its PM services are done in house.

Another report automatically notifies fleet managers when their trucks are 5,000 miles away from a scheduled PM. The report color-codes trucks in yellow when they are due for service in 2,500 miles. Trucks grouped in red are overdue, and managers are expected to get them to a shop within 24 hours.

For any report, users can hover above any metric with their mouse to see a description of how the data is defined.

“We don’t want report builders. We want true analysts,” Seibenhener said.

“You’ve got to be able to move, shoot and communicate,” Mead added. “If you are sitting there building reports and then have to analyze it, you just used up a whole bunch of time.”