Predictive analytics: The next step?
New technology promises accurate forecasting, allowing fleets to intervene with drivers before accidents occur
As the saying goes, “past performance is the best indicator of future results.” In the trucking industry, fleet safety managers increasingly are taking this adage to heart, employing any number of new technologies – including driver scorecarding and dashboard reporting – to hone in on poor-performing drivers in need of remedial training.
With new predictive analytics technology, fleet managers now can get even closer to forecasting future safety events. Vikas Jain, vice president and general manager for FleetRisk Advisors, a business unit of Qualcomm Enterprise Services, paid a visit to Commercial Carrier Journal’s editorial staff to discuss the company’s advancements in predictive analytics and remediation to improve accident mitigation.
While driver scorecarding and dashboard reporting show past and real-time driver performance, Jain says predictive analytics identifies root causes of driver stress – both internal and external – and anticipates the likelihood of a preventable accident occurring, a workers’ compensation claim or a driver leaving employment in the next 28 days. Identifying these at-risk drivers allows fleet managers to intervene.
“In-cab alerts are the last line of defense against an accident occurring,” says Jain, noting that safety personnel at larger fleets should think more broadly in terms of accident mitigation. “We would like to change the paradigm where you would anticipate the risk. By taking advanced action, you can actually prevent these events from even happening in many cases.”
FleetRisk Advisors aggregates more than 9,000 data points to determine the likelihood of specific drivers to have an accident before it occurs, allowing driver managers and dispatchers an opportunity to counsel and coach the driver beforehand.
“Drivers aren’t safe or unsafe – they are safe and unsafe,” says Jain. “There are times when they can become unsafe for a number of reasons. What we need to think about as an industry is why that happens. When we understand why, we can anticipate when a driver is going into a phase that will make him an unsafe driver and then take action before something happens.”
Currently, FleetRisk Advisors’ predictive analytics technology is suited to larger fleets (600-plus drivers) that can generate enough data needed to create accurate models. The company’s client list includes Maverick Transportation, Dupre Logistics, Roehl Transport and Averitt Express.
Refrigerated carrier C.R. England, which has 6,400 drivers, is also a client. It uses FleetRisk Advisors’ safety, workers’ compensation, recruiting and retention modules. Thom Pronk, corporate vice president of recruiting and training, says the technology supplements risk management by “giving us information that we never would have looked at before.”
“It has given us the ability to have conversations that we wouldn’t have had,” Pronk says. “Most of the risk behaviors can be traced to other things that are going on in a driver’s life.” Since using the platform, the Salt Lake City-based company has seen year-over-year accident rates improve.
But what if your company don’t have a driver work force in excess of 600 drivers? Jain says FleetRisk Advisors is working on a solution for medium-size fleets starting at about 200 drivers, and he expects the product to be ready by yearend.
FleetRisk Advisors offers predictive analytics and remediation for:
• Driver fatigue (Fatigue Model)
• Preventable accidents (Safety Model)
• Workers’ compensation claim (Workers Comp. Model)
• Good new driver recruit (Recruiting Model)
• Voluntary turnover (Retention Model)
The top 10% of drivers most likely to have an event within 28 days saw the following compared to the rest of the population of drivers in 2010-11:
• 60% fewer accidents
• 72% fewer workers’ comp. claims
• 81% less voluntary turnover
JEFF CRISSEY is Editor of Commercial Carrier Journal. E-mail email@example.com.