Trucking companies rely on more than rubber and fuel to operate; data from operations, sales, accounting and maintenance are vital to their services. Unfortunately, that data may be found anywhere from an AS400 or SQL server to a note scrawled on the back of a business card or an old bill of lading.
In addition, carriers since the early 1990s have used several generations of software and hardware, from the IBM mainframe to the Apple iPad, to collect data. Some trucking companies have several layers of “legacy” systems and older versions of software programs.
Enter the data cleanser. That’s the person who locates, sorts, updates and links data so it can be used by company executives in real-time decision making, to squeeze waste from operations and to find serious savings while improving customer service.
Motor carriers may not serve shippers as well as they could, or should, because the information they need is scattered to the four corners of their companies. A good portion of that data -- and possibly the equipment and systems used to process it -- is likely to be incompatible, incorrect, outdated or just impossible to find.
At U.S. Xpress Enterprises, “cleaning” data is the job of Tim Leonard, the chief technology officer the $1.4 billion truck-ing company recruited from $53 billion technology giant Dell. U.S. Xpress is known in trucking as an early adapter of new technologies, driven by co-founder and co-chairman Max Fuller. He brought Leonard, formerly a senior manager responsible for Dell’s data warehousing architecture, to the company’s Chattanooga headquarters last year to implement the latest in data-driven technology.
“When I first got here, they had a lot of data in silos in different areas,” Leonard said. “Some could have been on the AS400 server, some in other environments.” He also found a vast amount of data constantly streaming into the company’s computers. “Just with DriverTech, our in-cab information system, we get over 900 data elements with every pull,” he said, referring to each download of data from a remote truck.
Multiply that by more than 9,000 trucks and multiple downloads a day, and the scale of the company’s data architecture -- and Leonard’s challenge -- becomes clearer.
On top of that, much of the data was “dirty” -- incorrect, inconsistent or incomplete. Allowing such data to accumulate over time can corrupt entire systems.
“There’s no way we could have produced the types of dashboard and business intelligence reports we wanted because the data was just that dirty,” Leonard said. U.S. Xpress had “cleaned” its data about five years ago, he said, but unless data is cleaned repeatedly, it picks up “dirt” quickly and becomes impossible to use. Now data cleansing is a constant process at U.S. Xpress, running every five minutes.
“I recommended we not go any further until we took care of data quality,” Leonard said. He launched a data-quality initiative using applications from Informatica, a software vendor that provides data integration and management tools such as Informatica Data Explorer, or IDE, and Informatica Data Quality, or IDQ. Leonard used Informatica tools at Dell to support its multiyear enterprise business intelligence initiative.
At U.S. Xpress, he began with a pilot project to identify, collect and clean data associated with truck idling time. Much of the data came from truck engines, through DriverTech, and the idling time data itself was fairly clean, Leonard said. But it had to be related to other data before it would be useful.
“We had to create a small subset of information around the trucks and the drivers, and then cleanse that information,” Leonard said. “To produce the idle report, we needed to have the truck number. We needed to know who the driver was, who the fleet manager was, what region the truck operated in.”
There were a lot of elements to pull together, and it took some digging to find them all, he said. “DriverTech had the truck ID, for example, but didn’t identify the fleet manager accountable for that particular truck.”
The project took six weeks to complete and two weeks to test. “We had to run a lot of information through cleansing routines, and there were quality processes that we put in place,” Leonard said. “I knew I had to do something in six to eight weeks to prove that you can install and use a tool like Informatica and show the results. In trucking, you can talk all you want, but you’d better bring something to the table.”
What Leonard and his team brought to the table were reports backed by sound data estimated to save U.S. Xpress $6 million a year across its fleet of 9,200 trucks. The data quality software he installed paid for itself within three months.
“We contacted one fleet manager and told him he had a parked truck that had been idling for three days,” Leonard said. “He said, ‘If I had a truck idling for three days, I’d know about it.’ Then he walked out and there it was. Somebody parked it and left it running.”
From idling, the data quality initiative expanded to embrace all maintenance, operations and customer relations management. “We’ve been looking at silos of maintenance data,” Leonard said, and knocking those silos down to better manage the company’s most basic asset: its trucks. One maintenance “data-mart” and six executive dashboards replaced more than 400 reports, he said, allowing managers to quickly identify what service is due, which parts repeatedly fail, and more.
Cleaning and integrating the data gave managers a better idea of what they can get out of a truck, he said. “They’re getting a better understanding of how we use that truck, and starting to change their thoughts about how they manage that asset.”
That’s not an easy process, he said. “Sometimes people look at the data and say, ‘That’s wrong,’ but I’m not telling you what’s right or wrong, just what the data is telling you.”
Contact William B. Cassidy at firstname.lastname@example.org.