| Statistical Analysis of the Effectiveness of Electronic Stability Control (ESC) Systems – Final Report Report No. DOT HS 810 794 Jennifer N. Dang July 2007 ABSTRACT Electronic Stability Control (ESC) is a safety technology designed to enhance a vehicle’s stability and control in all driving situations. ESC first became available in the United States in 1997. Statistical analyses of 1997-2004 crash data from the Fatality Analysis Reporting System (FARS) and 1997-2003 crash data from the State data files estimate reductions with ESC for various types of crash involvements. - ESC reduced fatal run-off-road crashes by 36 percent for passenger cars and 70 percent for light trucks and vans (LTVs). The reductions are statistically significant.
- Police-reported run-off-road involvements were decreased by 45 percent in passenger cars and 72 percent in LTVs. The decreases are statistically significant.
- Fatal single-vehicle crashes that did not involve pedestrians, bicycles, and animals decreased (due to ESC) by 36 percent in passenger cars and 63 percent in LTVs. The decreases are statistically significant.
- ESC reduced police-reported single-vehicle crashes (excluding pedestrian, bicycle, animal crashes) by 26 percent for passenger cars and 48 percent for LTVs. The reductions are statistically significant.
- Rollover involvements in fatal crashes were decreased by 70 percent in passenger cars and 88 percent in LTVs. The decreases are statistically significant.
- Police-reported crashes involving rollovers were reduced by 64 percent in passenger cars and 85 percent in LTVs. The reductions are statistically significant.
- ESC reduced culpable fatal multi-vehicle crashes by 19 percent for passenger cars and 34 percent for LTVs. Only the reduction involving LTVs is statistically significant.
- Culpable involvements in police-reported multi-vehicle crashes were decreased by 13 percent in passenger cars and 16 percent in LTVs. The decreases are statistically significant.
- Overall, ESC reduced all fatal crashes by 14 percent for passenger cars and 28 percent for LTVs. Only the reduction in LTVs is statistically significant.
- Overall, police-reported crash involvements decreased by 8 percent in passenger cars and 10 percent in LTVs. The decreases are statistically significant.
EXECUTIVE SUMMARY Electronic Stability Control (ESC) is a safety system designed to recognize adverse driving conditions by 1) continuously measuring and evaluating the speed, the steering wheel angle, the yaw rate, and the lateral acceleration of a vehicle from various sensors and 2) using those measured data to compare a driver’s steering input with the vehicle’s actual motion. If an unstable situation is detected, then ESC automatically intervenes to assist the driver and stabilize the vehicle by applying the brakes to individual wheels as needed and possibly reducing engine torque. This technology is expected to reduce the number of crashes due to driver error and loss of control, because it has the potential to anticipate situations leading up to some crashes before they occur and the capability to automatically intervene to prevent them. A major benefit should be the reduction of single-vehicle crashes that involve losing control and running off the road. In September, 2004, the National Highway Traffic Safety Administration (NHTSA) issued an evaluation note on the Preliminary Results Analyzing the Effectiveness of Electronic Stability Control (ESC) Systems. The data suggested that ESC was highly effective in reducing single-vehicle run-off-road crashes. The study was based on Fatality Analysis Reporting System (FARS) data from calendar years 1997-2003 and crash data from five States from calendar years 1997-2002. The data were limited to mostly luxury vehicles because ESC first became available in 1997 in luxury vehicles such as Mercedes-Benz and BMW. NHTSA has now updated and modified its 2004 report, extending it to model year 1997-2004 vehicles – and to calendar year 2004 for the FARS analysis and calendar year 2003 for the State data analysis. Nevertheless, even as of 2004, a large proportion of the vehicles equipped with ESC were still luxury vehicles. Moreover, only passenger cars and SUVs had been equipped with ESC – no pickup trucks or minivans. The FARS database included fatal crash involvements from calendar years 1997 to 2004. The State databases included crash cases from California (2001-2003), Florida (1997-2003), Illinois (1997-2002), Kentucky (1997-2002), Missouri (1997-2003), Pennsylvania (1997-2001, 2003), and Wisconsin (1997-2003). The basic analytical approach was to estimate the reduction of crash involvements of the types that are most likely to have benefited from ESC – relative to a control group of other types of crashes where ESC is unlikely to have made a difference in the vehicle’s involvement. Crash involvements in which a vehicle 1) was stopped, parked, backing up, or entering/leaving a parking space prior to the crash, 2) traveled at a speed less than 10 mph, 3) was struck in the rear by another vehicle, or 4) was a non-culpable party in a multi-vehicle crash on a dry road, were considered the control group (non-relevant involvements) – because ESC would in almost all cases not have prevented the crash. The types of crash involvements where ESC would likely or at least possibly have an effect are: - Single-vehicle crashes in which a vehicle ran off the road and then hit a fixed object and/or rolled over.
- Involvements as a culpable party in a multi-vehicle crash on a dry or wet road.
- Collisions with pedestrians, bicycles, or animals.
The principal findings and conclusions of the statistical analyses are the following: RUN-OFF-ROAD CRASHES - ESC reduced involvements in all types of single-vehicle run-off-road crashes the following percentages:
Crash reduction by ESC (%) Cars LTVs Fatal crash involvements 36 70 Police-reported crash involvements 45 72 - All four of these reductions are statistically significant.
SINGLE-VEHICLE CRASH REDUCTION - ESC reduced all single-vehicle involvements (excluding pedestrian, bicycle, animal crashes) by the following percentages:
Crash reduction by ESC (%) Cars LTVs Fatal single-vehicle crash involvements 36 63 Police-reported single-vehicle crash involvements 26 48 - All four reductions are statistically significant.
ROLLOVER CRASHES - ESC was especially effective in preventing single-vehicle first-event rollovers.
Crash reduction by ESC (%) Cars LTVs Fatal rollovers 70 88 Police-reported rollovers 64 85 - All four of these reductions are statistically significant.
CULPABLE INVOLVEMENTS IN MULTI-VEHICLE CRASHES - ESC likely reduces involvements as a culpable party in multi-vehicle crashes.
Crash reduction by ESC (%) Cars LTVs Fatal culpable multi-vehicle crash involvements 19 34 Police-reported culpable multi-vehicle crash involvements 13 16 - Only the reduction in fatal crash involvements in passenger cars is not statistically significant.
COLLISIONS WITH PEDESTRIANS – BICYCLES – ANIMALS - There are no consistently significant results in either direction for crashes that involve pedestrians, bicycles, or animals
Crash reduction by ESC (%) Cars LTVs Fatal pedestrian, bicycle, animal crashes -36 -6 Police-reported pedestrian, bicycle, animal crashes 26 -11 - Only the reduction in police-reported crash involvements in passenger cars is statistically significant. We will continue to monitor the effect of ESC on this particular type of crash involvements in the future – because we do not have enough data (at the moment) for any conclusions.
OVERALL CRASH REDUCTION - ESC reduced all crash involvements by the following percentages:
Crash reduction by ESC (%) Cars LTVs All fatal crash involvements 14 28 All police-reported crash involvements 8 10 - Only the reduction in fatal crash involvements in passenger cars is not statistically significant.
4-CHANNEL VERSUS 2-CHANNEL ESC SYSTEMS IN PASSENGER CARS The passenger car sample includes certain make-models that had 2-channel ESC systems and others that had 4-channel systems. Separate analyses were performed to analyze the difference in effectiveness (if any) between 2-channel and 4-channel systems. We found: Greater fatal run-off-road reduction with 2-channel systems – but the larger observed fatality reduction with 2-channel systems is not statistically significantly different from the observed reduction with 4-channel systems. The reductions were most certainly influenced by the small samples. Larger reduction with 4-channel systems when police-reported crash involvements were included (as expected with larger samples). The larger reduction with 4-channel systems in all run-off-road involvements (mostly non-fatal crashes) is statistically significant. |
Print Version
