Truckload carrier Dupre’ Transport LLC, Alpharetta, Ga., today announced the results of the pilot project undertaken with FleetRisk Advisors to identify and remediate risk factors that may contribute to accidents.

The project consolidated historical data from Dupre’s internal systems into the FleetRisk’s Transportation Risk Analytics Center known as TRAC, and then running a series of analytical processes. The project team was able to determine unique risk signatures, which will assist Dupre’s senior management in developing action plans to improve safety and increase profitably. Based on these findings, Dupre’ Transport has elected to extend its relationship with FleetRisk Advisors and has entered into a seven-year strategic agreement for ongoing analysis of its operations.
The overall project consisted of three analytical processes:
Aggregation Analytics This process provided new and interesting results, including:
* Dupre’s newest drivers as a group actually are less likely to have a loss causing incident than more tenured drivers.
* Additional training such as HazMat recertification dramatically reduces incident frequency.
Risk Signature Analytics This process provided insight into the relevancy and predictive nature of key data elements. For example:
* Prior driving experience and especially the type of commodities hauled have a high level of predictability with regard to accidents/incidents.
* Drivers with hard braking and/or overspeeding events were twice as likely to be involved in a loss causing accident.
Predictive Modeling and Pattern Recognition The third and most complex process used seven years of historical driver and accident information. A predictive model was built using the first six years of data and then “blind tested” using the seventh year’s data. The modeling process identified five key data elements that proved to be highly predictive of future accidents and incidents. Based on those data elements, the current driver pool was ranked in order from highest propensity for an accident/incident to the lowest. The validation test showed that the group of driver ranked as “most at risk” actually had over three times the number of accidents and incidents as the group ranked as “least at risk”.
Dupre plans to implement several predictive models with FleetRisk Advisors; a driver scoring model, a new applicant scoring model, a dispatch risk-score model, and a customer profitability model which takes safety into account.
“FleetRisk has been able to give us a window into our operations we never had before,” said Reggie Dupre’, president and CEO of Dupre’ Transport.
“This project has shown how pattern recognition can integrate existing operational data with in-vehicle monitoring, driver aptitude testing, fatigue management and other systems to provide exception based information to avoid accidents, mitigate risks, and save money,” said Sam Wilkes, CEO of FleetRisk Advisors.
One of the driving forces behind FleetRisk’s service is a suite of analytical tools, including pattern recognition and predictive modeling powered by Valen Technologies. This technology analyzes information about the fleet, including data from telematics systems, searching for correlations and patterns that become predictive risk signatures.
For more information, go to www.fleetriskadvisors.com.