Written in the Miami, Oklahoma Accident Report (NTSB/HAR-10/02 PB2010-916202)
In its 2001 special report on technology for the prevention of rear-end crashes,187 the NTSB reported that, in 1999, the DOT had begun operational testing of ACC systems and FCWSs for cars and trucks. The NTSB also reported that rear-end collisions accounted for 1.8 million crashes in 1999, including 1,923 fatal crashes. Of the fatal crashes, 770 involved commercial vehicles (trucks weighing more than 10,000 pounds and motorcoaches). Thus, CMVs were involved in 40 percent of the fatal rear-end crashes, even though they accounted for only 3 percent of vehicles and 7 percent of miles traveled. Although the NTSB has acknowledged that an FCWS is not intended to replace driver vigilance, such a system can aid drivers when they are distracted or fatigued, or when their attention is concentrated on something other than the road ahead. The NTSB concluded that accident statistics and the investigation findings indicate that accident consequences are more severe when commercial vehicles are involved in rear-end collisions and that the public can benefit from technology designed to help prevent such collisions. As a result, in its special report, the NTSB asked the DOT to implement Safety Recommendations H-01-06 and -07. Safety Recommendation H-01-6 is on the NTSB Most Wanted List of Transportation Safety Improvements in the issue area “Prevent Collisions by Using Enhanced Vehicle Safety Technology.” Deployment of vehicle collision avoidance technology has been on the Most Wanted List since November 2007.
In the Miami accident, an FCWS alert could have drawn the accident driver’s attention to the hazard ahead, which was the slowing traffic. The truck-tractor semitrailer was traveling about 103 feet per second. With the maximum available warning detection distance of 350 feet provided by an FCWS, the Miami accident driver would have received a warning from the system about 3.40 seconds before striking the rear of the slowly moving traffic queue. Within this 3.40-second warning period, the driver would have to have (1) been effectively alerted; (2) comprehended the severity of the alert and the situation ahead; and (3) mechanically executed a reaction, including moving his foot from its rest location (the cruise control was engaged) and placing it on the brake and applying maximum (emergency) braking immediately. If any time (and distance) had remained from the 3.40 seconds after (1) through (3) above, it could have gone toward slowing the vehicle or enabling the driver to take an evasive action to mitigate the impact force of the tractor semitrailer upon the passenger vehicles.189
It should be noted that, for an alert driver,190 the average projected reaction time to an unexpected situation can range from 0.75 second to about 2.50 seconds for a 90th percentile driver.191,192 Research supports that, in the middle of this range, drivers have a perception reaction time to a common but unexpected stimulus (such as the unanticipated brake lights of a car ahead) of about 1.25 seconds.193 Given a reasonably clear and straightforward situation, most drivers will respond within 1.50 seconds of the first appearance of an object or condition of concern;194 they will react to a surprise event (such as an object moving unexpectedly into the vehicle’s path) in 1.50 to 1.75 seconds.195
Some FCWSs are equipped with ACC,196 which uses the same detection technology as the FCWS to adjust or disengage the conventional cruise control when it is in use. An active braking system that can automatically apply the foundation brakes197 of the vehicle is also an available technology. If a collision is deemed imminent, an FCWS with active braking does not wait for the driver to react; in such a critical situation, braking is applied automatically to reduce the severity of the impending collision. Once active braking is initiated to mitigate the accident
(not when initiated to slow the vehicle to maintain following distance, such as with the ACC), it also may be referred to as “collision mitigation braking” or CMB. When these technologies are bundled together, they are often referred to as “collision mitigation systems.” If the Miami accident truck had been equipped with an FCWS that included an active braking system, the driver’s reaction time would not have been a factor—only the brake lag time would have contributed to the distance traveled before maximum braking was achieved.
An FCWS alone or bundled with ACC and active braking could have significantly affected the outcome of the Miami accident, depending on a number of factors, including the point at which the system detected the Land Rover ahead of it. The unloaded truck-tractor semitrailer had a gross weight of 40,400 pounds (a loaded truck-tractor semitrailer can weigh up to 80,000 pounds). Most cars weigh less than 4,000 pounds. Thus, when a commercial truck strikes a passenger car in the rear, the large difference in mass between the vehicles means that this impact most likely will not bring the heavy truck to a stop or even slow it appreciably; consequently, the impact itself does relatively little to keep the truck from continuing to move and to involve more vehicles.198 An FCWS can reduce the risk of these rear-end crashes by identifying fast-closing speed situations and providing the driver with additional time to react. It should be noted that ACC systems are designed to maintain a predetermined199 following interval behind another vehicle, thereby providing more time to resolve driving conflicts to reduce the probability of a rear-end collision.200
An FCWS with active braking can begin to decelerate a vehicle automatically, having the added benefit either of reducing the speed of the vehicle if the driver does not intervene or of supplementing deceleration before the driver applies braking. Active braking systems, such as the Bendix Wingman ACB (active braking with cruise control) and the Meritor WABCO OnGuard, do not apply the foundation brakes at the full emergency brake application level that a driver can.
To illustrate some possible scenarios for this accident under different circumstances, the NTSB worked with several FCWS manufacturers201 and developed some potential outcomes had the accident truck been equipped with an FCWS alone or bundled with ACC and/or active braking. Three of the possible outcomes are presented below. NTSB investigators were unable to determine the speed of the Land Rover just prior to its being struck by the Volvo; therefore, the first two of the following scenarios are based on witness interviews indicating either that the struck vehicles were stopped in traffic or that the Land Rover was moving slowly (just over 10 mph). The third instance is the “best case” scenario, in which the Land Rover was just beginning to decelerate from the posted speed limit of 75 mph when the Volvo FCWS detected it at 70 mph, with the Volvo 350 feet or more behind it. All calculations in the scenarios and tables below considered a roadway coefficient of 0.65 g deceleration for the Volvo and an initial truck speed of about 70 mph. They also assumed the postaccident inspection condition of the brakes, which were within adjustment limits, on the truck-tractor semitrailer. An air brake lag time of 0.50 second was used, in addition to the driver perception reaction times of 2.50, 1.50, and 0.75 seconds. The term “distance to decelerate” used in the tables below is the distance between the accident truck-tractor and the Land Rover when the truck driver receives the first FCWS alert. The “warning time” is the time the truck driver would have between the first FCWS alert and the estimated impact.
Scenario 1—FCWS and Land Rover Stopped. Had the Land Rover, the first vehicle struck by the Volvo truck, been stopped (stationary) in the traffic queue, an FCWS on the Volvo could have detected it at either 308 or 350 feet,202 calculated the closing distance, and sounded an audible alert. Table 7 shows the reductions in impact speeds possible, had the Volvo truck driver perceived the meaning of the alert and reacted, given the 0.75- to 2.50-second range of driver perception reaction time to the FCWS warning. This table shows the possible outcomes using FCWS alone, without the added benefit of ACC or active bra
The Volvo driver, although he was fatigued, was not incapacitated, and had he received an alert warning from an FCWS, he might have reacted with emergency braking. The accident truck could not have slowed down tremendously, given the assumption of a 2.50-second perception reaction time and the stopped traffic ahead. Under such circumstances, it can be estimated that the impact speed range would be 70 to 64 mph. If the FCWS alert had immediately redirected the driver’s attention to the traffic ahead, and the driver had reacted very quickly, faster reaction times of 1.50 and 0.75 second would have reduced the impact speed to a range of 56 to 39 mph. Although 56- and 39-mph impacts are significant, they are less severe than a 70-mph impact. In addition, at the lower speeds, the Volvo driver might even have been able to take evasive steering action to avoid or mitigate the accident. The driver could have attempted an evasive maneuver, such as steering to the right, onto the roadway’s paved shoulder, or even off the road and onto the grassy right-hand right-of-way, to prevent striking the passenger vehicles.
Scenario 2—FCWS and Land Rover Moving Slowly. Had the Land Rover been moving slowly in traffic at 10 mph, the truck-tractor combination unit would have gained an additional 44 to 50 feet of distance over which to decelerate in this scenario, depending on the warning and perception reaction times used. The radar-based FCWSs would have detected the Land Rover at a range of 350 feet: one system would have emitted the audible alert at 350 feet, while another would have calculated the closing distance and sounded an audible alert at approximately 318 feet in closing distance. The camera-based system would have detected the slowly moving Land Rover at a following distance period of 3.00 seconds, which equates to 308 feet, and would have emitted an alert at this distance.
Table 8 shows that, had the traffic ahead been moving slowly, affording the Volvo truck a longer time and greater distance over which to decelerate, the impact speed of the truck could have been reduced to 38 mph under the most conservative reaction and warning time assumptions. Assuming a quicker driver reaction time of 1.50 seconds, the Volvo’s impact speed could have been reduced to a range of between 24 and 14 mph. Given a driver reaction time of 0.75 second, the impact speed might have been reduced to as low as 9 mph, or the impact might even have been avoided.
One manufacturer indicated that with a bundled system on the Volvo truck, consisting of an FCWS with ACC and active braking, the driver could have brought the vehicle to a stop if he had applied 0.60 g emergency braking approximately 2.00 seconds after the active braking system engaged. In this scenario, the active braking itself might have alerted the driver to the impending hazard and caused him to initiate an appropriate response. According to the manufacturer, even if the driver had not initiated any emergency braking but the Volvo had been so equipped, this system might have been able to initiate CMB and slow the Volvo to an impact speed range of 48 to 53 mph without any driver action.
Scenario 3—FCWS With Bundled System and Land Rover Beginning to Decelerate From 75 mph When the FCWS Detects it at 70 mph. Both of the scenarios described above assume that the Volvo truck-tractor was 350 feet behind the Land Rover (or any other vehicle) when the FCWS detected it as stopped or slow-moving traffic. If, instead, both vehicles were traveling about 70 mph when the truck’s FCWS detected the Land Rover—with at least 350 feet of separation distance—and the FCWS had been tracking the Land Rover when it began to slow in response to the traffic queue, this could have affected the accident outcome significantly.
In this case, if the Volvo truck had been equipped with an FCWS with ACC and active braking, the system would have automatically slowed the Volvo to a preset safe following distance (one manufacturer’s default setting is 3.60 seconds) without driver input. Further, once the system detected that the vehicle ahead was continuing to slow, the Volvo with the FCWS, ACC, and active braking would have maintained the 3.60-second following distance by continuously slowing. When the Land Rover reached 0 mph, the Volvo truck-tractor semitrailer would also have slowed to 0 mph at a distance of 32 feet behind the Land Rover, thus entirely preventing the accident. The above “best case” scenario illustrates what might have been possible in the Miami accident with a vehicle equipped with an FCWS with active braking; under these very specific circumstances, such a system could have prevented an accident without any driver input.203
As discussed earlier, the Volvo’s impact speed generated tremendous kinetic energy, which was dissipated when it collided with the slower moving passenger vehicles, causing them catastrophic damage. Kinetic energy is the mathematical expression of the truck’s maximum ability to do damage.204,205 Because kinetic energy is proportional to the square of the vehicle speed, the energy of the impacting vehicle and its ability to do damage decline quickly as speed is reduced. Table 9 below shows the amount of kinetic energy that the accident Volvo had at about 70 mph, when it struck the passenger vehicles, as well as the amount it would have had with the incremental reduction in speed provided either by an FCWS alone or by an FCWS with a bundled system, as described above. A reduction in speed from about 70 to 50 mph would have cut the kinetic energy of the impacting heavy commercial vehicle in half. Further reducing the impact speed to 39 mph would have caused an energy reduction of nearly 70 percent. The scenario of the FCWS system bundled with ACC and active braking, without any input from the driver, could have resulted in a reduction in speed from about 70 to 39 mph at impact. (See table 9.)
Depending on variables (such as the speed and distance of the vehicles ahead of the Volvo truck), even a bundled system might not have provided the fatigued Volvo driver sufficient time to react to the warning, brake the vehicle, and prevent the accident. However, it could have provided enough time for him to react, brake, and mitigate the severity of the accident or perhaps to avoid the collision through steering inputs.
It might not have been possible to bring the heavy Volvo truck-tractor semitrailer to a complete stop with FCWS and related technologies before any collision occurred. However, as can be seen in table 9, the slower the truck had been traveling at impact, the lower the kinetic energy involved in the accident and the less severe the damage to the struck passenger vehicles would have been. This scenario most likely would have resulted in less severe injuries to the occupants of those vehicles. In fact, if the Volvo truck-tractor had been equipped with an FCWS bundled with ACC and active braking, assuming that scenario 3 circumstances had existed in the seconds before the accident, it is possible that the system could have entirely prevented the accident. Therefore, the NTSB concludes that an FCWS with ACC and active braking would have provided the driver with the best opportunity to prevent, or reduce the severity of, the truck-tractor semitrailer’s impact with the passenger vehicles in the traffic queue.
The NTSB considers that installing new technologies in CMVs—such as FCWSs, ACC, active braking, and ESC—has the potential to reduce accidents substantially. Following the investigation of an October 2005 accident in which five people were killed when a motorcoach collided with an overturned truck-tractor semitrailer combination unit on Interstate 94 near Osseo, Wisconsin,206 the NTSB issued Safety Recommendation H-08-15 to NHTSA.
Since February 26, 2010, Safety Recommendation H-08-15 has been “Open—Acceptable Response.” Also in the Osseo report, the NTSB reiterated Safety Recommendations H-01-6 and -7 to NHTSA.
In a letter dated June 4, 2009, NHTSA responded to these NTSB recommendations by providing an update on its current projects evaluating the application of various technologies for commercial trucks and motorcoaches. NHTSA is conducting a test track evaluation of commercially available CMB systems and has indicated that an initial evaluation of their performance capabilities will be completed in 2010. A NHTSA project to evaluate the potential safety benefits of active braking systems is expected to be completed in 2011. Based on these reports of progress from NHTSA, Safety Recommendations H-01-6 and -7 were classified “Open—Acceptable Response.”
Due to their high mileage exposure207 and the severity of crashes involving them, combination-unit trucks have the highest crash cost per vehicle over the operational life of the vehicle; therefore, FCWSs may provide a relatively higher safety benefit for this class of trucks.208 However, government and industry entities are still conducting operational testing and encouraging voluntary implementation of FCWSs. Although the work being done by private industry and the government is encouraging, the slow pace of testing and standards development and the limited deployment of FCWSs in commercial vehicles are cause for concern, given the large number of rear-end collisions and the high rate of fatalities that result when commercial vehicles are involved.
For years, the NTSB has been advocating the implementation of in-vehicle systems that enhance the safety of heavy vehicles, both by mitigating accident severity and preventing accidents altogether. Safety benefits are often not the result of one system on its own; more often, it is the synergy of systems working together that can prevent and mitigate a larger percentage of accidents, resulting in the greatest reduction of highway injuries and fatalities. Although FCWS use within a heavy vehicle is crucial to provide warning of an impending collision, integrating this safety system with related technologies would provide even greater opportunity for preventing accidents, as well as for reducing the severity and frequency of rear-end accidents. The NTSB considers that FCWSs have great promise and that the added feature of active braking increases their potential for preventing accidents. However, the pace of NHTSA’s progress in this vital area has been too slow. Because NHTSA is still evaluating these systems and is not yet near rulemaking that would require them to be used in commercial vehicles, the NTSB reiterates Safety Recommendations H-01-6 and -7 and H-08-15. Further, although the NTSB acknowledges that NHTSA has made some progress in conducting research in this area, due to the lack of timely completion of the recommended actions, Safety Recommendations H-01-6 and -7 are reclassified OPEN – UNACCEPTABLE RESPONSE. The status of Safety Recommendation H-08-15 remains “Open – Acceptable Response.”