Over the next several years, it's expected that some 50-100 new sensors will be added to trucks beyond those already there, and a truck could soon reach a point of producing a terabyte of data a month. That's quite a sea of ones and zeros to wade through for fleet managers, and if you're not savvy with data analytics yet, that first dip of your toe in the water will only get more challenging.
Those points came up at the recent Omnitracs Outlook 2016 user conference in Dallas. Vince Fiore, manager of consulting services for Roadnet Technologies, and Craig Powell, senior market analyst and product manager for Roadnet, shared some advice for fleets and trucking companies looking to get started or assess where they're at with data analytics — and noted a few pitfalls to avoid.
1. Understand your company vision and objectives and identify your business drivers. That'll be important in helping target improvements with data analytics, Fiore and Powell said, but even more so in translating information gathering and analysis to taking real action steps.
2. You'll need executive sponsorship/buy-in with data analytics to improve your operations "or it's not going to happen," Fiore told Fleet Owner.
3. Be sure your data points are clean for the information to be meaningful. "That's always an issue that we're dealing with," noted Fiore.
4. You'll need to educate personnel involved on the importance of data metrics, Fiore added, "because you'll also need to get buy-in from your people."
5. Data analytics "needs to be embedded in your company operations" and become like second nature, Fiore contended: "You want it at some point to be something you're not even thinking about."
6. With the right foundation for success with data analytics — note the above — the next step is to create some primary KPIs, or key performance indicators.
Fiore also stressed the importance of taking the right steps in the right sequence when approaching data analytics rather than "jumping too far ahead," noting that many fleets he's worked with end up doing isolated elements of the steps they'll need to be successful. And on that note, Powell pointed out that fleets often put the cart before the horse, essentially.
Fleets looking to use telematics data on driver behavior, for example, may wish to set goals and targets. "But you don't want to do that if your data isn't clean and people don't understand why you're putting targets in front of them," he said.
"A lot of people want to jump right to numbers and targets, numbers and targets," Powell continued, "but when you do that, it usually isn't very successful because you haven't done all the preliminary work to get it accepted and understood inside the organization."
So across the industry, where is trucking when it comes to data analytics? "It goes from elementary school all the way up to PhD," Powell quipped.
"What I see is I think people track numbers a lot. What they have a hard time doing is converting looking at numbers into actually driving business and reducing costs or increasing efficiency," he added.
Fiore noted that Omnitracs is placing more of an emphasis on its analytics consulting team, and said services of that kind can help fleets get started more effectively and achieving results.
"It is such a big process. I think that's why you'll see so many [fleets and fleet managers] doing isolated steps," he suggested. "You often need help. It's like when you're building a house — you might have architects, contractors, builders and finishers. This is a big process, and sometimes people don't realize how large it is."
Likewise, Powell said fleets need to find ways to hone down their focus when doing analytics, and they're often spending too much time on the wrong data. "To squeeze money out of your operations and reduce costs a little at a time, what you really need to do is find those niche areas," he contends. "You need to find the 20% of your business that is outside the norm and does not meet your standards that you actually can improve on."
That's where data tools like Omnitracs Roadnet Insight can help sift through your fleet data to display, summarize or otherwise make it more readily usable for managers, he said. "You need tools to help you find problems at a real low level, and then turn around and highlight them at a high level to let you know, 'Hey, my average cost per mile on the East Coast is looking pretty good, but I still have problems in Atlanta. And in Atlanta, I still have problems in these three routes,'" Powell told Fleet Owner.
"There's just tons of data out there," he added. "There's a lot going on. And you cannot spend hours a day looking at the part of the business that is running normally."
The idea is to find those fringe areas not only in terms of low but high performance, according to Powell, since high-performing operations segments can serve as examples and could be replicated.
"You need tools to help find those outliers so you can concentrate on those areas that need attention," he noted. "Time management is a huge thing."