Driver Safety Scorecards Help You Stay Ahead of CSA

The Federal Motor Carrier Safety Administration's double whammy of proposed new hours of service (HOS) rules and the emerging CSA 2010 driver and carrier safety rankings is prompting many transportation departments to find better ways to manage safety compliance and driver safety. Most private fleets already use on board computer systems (OBCs) that help drivers maintain HOS logs and track driver behavior. Any carrier with a government-issued DOT number may also access the government's motor carrier Safety Measurement System (SMS) to examine individual driver safety scores as well as the carrier's rolled up score. So, there is plenty of safety data to work with. The challenge lies in leveraging that data to make decisions on a daily basis—decisions that impact your carrier ranking and compliance performance over time.

Safety scorecards help transportation professionals simplify the data challenge. Using a good score card minimizes the time required to analyze data, draw conclusions, and execute strategies for changing driver behavior and adjusting key processes. Instead of looking at raw numbers, a scorecard approach quantifies several data points and then assigns a numeric and/or visual value (e.g. color code) that helps quickly and efficiently rank performance. In the case of drivers being ranked for safety, you might take a measurable indicator of safety, such as speeding incidents and heavy braking, and assign each a score.

Example of a Safety Scorecard

Here then is how you might craft a simple scorecard.

Every time an OBC logs a highway speed 5 miles over the posted speed limit each speeding incident could be count as a "1." Likewise, every rapid deceleration might be counted as a "1". If a driver had one speeding incident and one rapid deceleration incident the driver's score would be 2. If another driver had one speeding incidents and two rapid deceleration incident in the same time period, his score would be a three. In this case, sorting on the safety score, the first driver would have a lower score.

To understand these scores, you have to understand how the scores were derived. In this instance, you would have to know that the lower score is the indicator of better safety performance. If you had 80 drivers to evaluate, and you sorted on this simple score from low score to high, your best performing drivers would be at the top of the list. To have confidence in the score, you also have to have confidence that the original data is accurate and that the formula for coming up with the score has its own persuasive logic. If you aren't persuaded that five miles over the limit is predictor of accidents, then you might prefer to change the calculation to ten miles over.

Let's say in our example, rapid decelerations count are twice as likely to indicate a safety risk, then you might weight it more heavily. Let's say you multiply the deceleration score by 2 before adding it to the speed score.

Scorecard

Driver #1 variables: speed = 1 , deceleration = 1. [1 + 1 = 2 Safety Score]

Driver #2 variables: speed = 1 , deceleration = 2. [1 + 2 = 3 Safety Score]

Revised scorecard (Same variables but factors weighted differently)

Driver #1 variables: speed = 1 , deceleration = 1. [1 + (1*2) = 3 Safety Score]

Driver #2 variables: speed = 1 , deceleration = 2. [1 + (2*2) = 5 Safety Score]

Thus, using the revised scorecard, Driver #2's performance is slightly worse compared to driver #1.

Good Scorecards Should Be Interactive

The problem with the government's SMS website is that the safety performance score is presented in tabular, text-heavy form that is not easy to evaluate or manipulate for many folks. A good scorecard should include controls to quickly and easily change date ranges to do some trend analysis. So, for our above example, is the score for one day, one week or one month? Maybe Driver #1 is the best for last week, but the worst over the past month. Maybe driver #2 can be placed in a secondary category for further analysis, such as short haul vs. long haul. There are nearly as many ways to aggregate data into simple views as there are the raw numbers behind the view. If your scorecard exists only as a static template that you can't manipulate, it has limited value for fact-based decision making. If you are going to hold drivers accountable, even to the extent of letting the worst-performing driver go, then you better have confidence in your scorecard analysis, scenario testing, and reporting. Drivers and senior executives both should be able to use the scorecard and draw their own conclusions intuitively.

You should also be able to drill down on aggregate score and see the different factors that comprise it. For driver safety performance, you should be able to click through to get any number of facts about the driver you need to reach a solid conclusion about next steps.

Good Scorecards Should Provide Context

What does a given number mean? Is it accurate? Is it good or bad...relative to what? What is the context? Perhaps you need to evaluate the score relative to industry or government benchmarks, or against last year's performance. What is the trend and how can you measure success? How well does your scorecard on safety help you visualize performance using charts, and sophisticated means such as bubble plots, sparklines, and embedded bullet graphs? A pretty gauge that wastes space just to display a number is less valuable than a compact graphic that provides context and the ability to drill down into the data or roll up findings into a summary. You should also be able to change calculations, filter the data, and modify your charts and graphs on the fly (ad hoc) without going back to a software vendor for an expensive custom one-off report.

Doing Something about HOS

OBC's help log HOS, but they don't necessarily show you what to do. With Syntelic's Resource Planning tool, when transportation puts together a daily dispatch plan, HOS for every driver is available when determining optimal route assignments. And as HOS rules change, the system lets you change those numbers yourself, and see the effect that has on the comprehensive resource picture, including master routes, bid packages, and available equipment. You can plan routes two days out and see at a glance how those plans impact each driver's availability to run a given route and stay compliant.

The Syntelic Difference

By design, Syntelic acts as one point of entry for all of your safety analysis. Instead of being limited to a set of standard reports or dashboards, Syntelic provides user-generated drop-and-drag dashboards. The format of each view can be changed by the user, who can select a graph, chart, or data table. Elements of those data visualizations are also under the user's control, and the calculations, such as how a given score is determined, is likewise under the user's control. Syntelic integrates multiple streams of safety information, including HOS logs, engine metrics from the OBCs, and if needed, ways to import other files such as human resources records on drivers, equipment maintenance logs, and benchmarking to provide context such as findings from studies like the one recently conducted by the American Transportation Research Institute. Syntelic can even bring in the government's CSA data to let you better manage that data outside of having to log into the SMS website, with its limitations for manipulating the safety data embedded there.

Articles