SafeStat Explained
SafeStat uses data from five sources: the National Governors Association database of reportable accidents collected by state and local police; safety audits of companies by federal auditors; results of enforcement cases; results of roadside inspections; and the motor carrier census form filled out by carriers applying for their Department of Transportation registration numbers.
But the system isn’t just some black box that spits out a list of who should get a visit from DOT’s safety specialists. At its heart is a highly detailed algorithm that assigns weights to the different pieces of information.
In scoring a carrier’s level of safety, all the data are not counted equally. Instead, the weighting system places greater or lesser value on different types of information. For example, the age of the data makes a difference, with data from the last six months counting most, followed by data seven to 18 months old counting less and data 19 to 30 months old counting the least. Data older than 2½ years old is not considered at all.
Weighed data are used to calculate measures. These measures summarize some aspect of a carrier’s safety performance and represent it as a number. The measures are then ranked relative to the rest of the industry to create percentile indicators.
A percentile of 75 in an indicator means 75% of carriers performed better than the company and only 25% performed worse.
Indicators are grouped into four categories identified as safety evaluation areas (SAEs). These areas are: accidents, drivers, vehicles and safety management.
For the full story, see the Sept. 6 print edition of Transport Topics. Subscribe today.