Predictive Modeling & PRiMASTM

Unmatched in its capabilities, FleetRisk Advisors develops proprietary models for clients that identify (predicts) drivers "most at risk" to have an accident in the next period (month, week or day). These predictions are coupled with an intelligent intervention system that positions you to prevent accidents from occurring.

Tom Voelkel, President and COO of Dupre Transport is quoted in a recent article as saying... "Predictive modeling helps us see the future today and gives us the opportunity to create the future that we want tomorrow..." "This is a major factor in moving us closer to our vision of being the safest transportation and logistics company in North America,"              Predictive Modeling Pyramid

 

Driver Safety – the driver safety model predicts the accident risk characteristics of each driver. The model produces a monthly score for each driver, ranking the drivers from most at risk to least at risk to have an accident. Clients typically create a Management Action Plan for the top ten percent (10%) of drivers identified to be “most at risk”. Clients vary the number of driver-manager interventions depending on their fleet accident rate and other factors.

 

 Driver Fatigue – this model is intended for those drivers required to keep a logbook in accordance with the US Department of Transportation Hours of Service regulations. FleetRisk accesses data from a driver’s duty status from their logbook and produces fatigue scores for all drivers. Drivers with high fatigue scores may be identified for a fatigue Management Action Plan.

 

 

Driver Worker’s Compensation – the worker’s compensations model produces a monthly score for each driver currently employed and ranks them from most at risk to file a worker’s compensation claim. Clients generally establish a threshold for managers to create a Management Action Plan for managers to work with a fixed percentage of drivers identified as most at risk.

 

 

Driver Retention – the retention model will produce a monthly score for each driver currently employed and ranks them from most at risk to voluntarily terminate employment to least at risk to voluntarily terminate employment. Client managers are expected to develop Management Action Plans to retain the “best” drivers who are most likely to terminate their employment. The Retention Model includes a wide data set ranging from driver performance metrics (miles, service, MPG, etc.) to payroll information (gross pay, advances, net pay) to load assignments.

 

 

Driver Selection – this model takes data from the client’s recruiting process (including any on-line application systems) as well as data from other sources to create a predictive model that identifies which driver applicants will stay with the company for a minimum period of time. Alternatively, the model can be designed to identify applicants likely to be high performing, safe, long-term drivers.