Daily Rules, Proposed Rules, and Notices of the Federal Government
On October 27, 2010, the agency received a petition for rulemaking from William B. Trescott of Bay City, Texas, requesting that FMVSS No. 121, Air Brake Systems, either be vacated entirely or amended to require one of two options regarding antilock brake systems that are required for air-braked vehicles. The first option would be to require automatic deactivation of the antilock brake system (ABS) when vehicles are travelling at speeds faster than 55 mph, and the second option would be to require an ABS deactivation switch to allow the driver to disable the ABS.The petition cited data from a recent NHTSA report, “The Effectiveness of ABS in Heavy Truck Tractors and Trailers,”
The petition cited specific roadway type, speed, and locality data that are contained in the report as follows. Table 17,
The petition stated that an unintended consequence of preventing jackknife crashes through the use of ABS is that incompetent drivers, who prior to the introduction of ABS would have been fired for the occurrence of a jackknife, were instead being retained and subsequently their continued driving resulted in increases in other types of crashes. The petition cited a 29 percent increase in two-vehicle rear end crashes on wet or icy roads with the truck as the striking vehicle, from Table 4,
The agency's study on the effectiveness of ABS on tractors and trailers included a statistical analysis of crash data from seven states for fatal and non-fatal crashes that occurred between 1998 and 2007 (data for all of these years were not used or were not available for every state), and from the Fatality Analysis Reporting System (FARS) for fatal crashes that occurred between 1998 and 2008 from all fifty states. All states provided the vehicle identification number (VIN) or the model year data for the tractors so that the model year of the tractor could be determined, while only two states for which trailer ABS was evaluated (Florida and North Carolina) included the VIN or the model year for the trailers. For all of the crashes, the data were limited to a tractor towing one trailer; thus tractors not towing a trailer (bobtail tractors) or tractors towing multiple trailers were not included in the analysis. Tractors of model year 1998 or newer were assumed to have ABS while those of model year 1996 or older were assumed not to have ABS. Model year 1997 was excluded since the ABS requirements in FMVSS No. 121 became effective on March 1, 1997, and therefore a model year 1997 tractor may or may not have been equipped with ABS. Similarly, trailers of model year 1999 or newer were assumed to have ABS, while those of model year 1997 or older were assumed not to have ABS, and model year 1998 trailers were excluded from the analysis, since the trailer ABS requirements became effective on March 1, 1998.
Limitations of the study included the overall small vehicle population for tractor-trailers (compared to light vehicles for which there are many more vehicles on the road) and the limited amount of crash data from the seven-state sample (27,777 total crashes). Additionally, all model years of vehicles prior to the ABS effective date were assumed not to have ABS, which did not account for an unknown number of vehicles that were voluntarily equipped with ABS prior to the effective date. Also, there was no way to discern whether the vehicles equipped with ABS had been properly maintained so that the ABS was functional at the time of the crash; both of these factors would result in underestimation of the ABS effectiveness. As described above, only two states had information on trailer model year, so the main focus of the analysis was on the effectiveness of tractor ABS.
The crashes, in which tractor-trailers were involved in either single vehicle crashes or multiple vehicle crashes, were divided into control and response groups that both contained tractors and trailers with and without ABS. The crash types for the control group were those in which ABS should not have been influential in the crash outcome, including crash involved tractor-trailers that were moving slowly, parking or unparking, backing up, impacted in the rear, etc. The crash types for the response group were those in which ABS should have been influential either by helping the driver to maintain control of the vehicle or by contributing to improved stopping distance. Response group single vehicle crash types included run-off-road collisions with fixed objects; collisions with animals, pedestrians, or bicycles; jackknife crashes, etc. Response group multi-vehicle crashes included those in which the truck was the striking vehicle in rear-end crashes or the truck was the at-fault vehicle in any other type of crash involving other vehicles. Differences in control group and response group crashes were used to determine ABS effectiveness as evidenced by reductions or increases in crashes among the response group, and statistical measures were provided to determine the statistical significance of the results.
The primary findings of the analysis are summarized as follows:
• The best estimate of a reduction by ABS on the tractor unit in all levels of police-reported crashes for air-braked tractor-trailers is three percent, based on crash data from seven states and controlling for the age of the tractor at the time of the crash. This represents a statistically significant six percent reduction in crashes in which ABS is assumed to be potentially influential, relative to a control group, of about the same number of crashes, in which ABS was likely to be irrelevant.
• In fatal crashes, there was a non-significant two percent reduction in crash involvement, resulting from a four percent reduction in crashes in which ABS should have been potentially influential. External factors of roadway urbanization and speed, and ambient lighting, were accounted for in the final estimates.
• Among the types of crashes that ABS influences, there is a large reduction in jackknife crashes, off-road truck rollovers, and at-fault involvements in crashes with other vehicles, except in rear-end crashes. Counteracting was an increase in the number of involvements in crashes with animals, pedestrians, or bicyclists and, only in fatal crashes, two-vehicle rear-end crashes with the truck as the striking vehicle.
The first stage of the analysis considered ABS on both the tractors and the trailers. For the Florida data, the reduction in response group crashes was a statistically significant 14 percent for ABS-equipped tractors when towing either ABS-equipped trailers or non-ABS-equipped trailers.
When the Florida data were limited to wet roadways (with the road surface coded as wet, slippery, or icy), the reductions in crashes for ABS tractors were even higher: 26 percent when operated with non-ABS-equipped trailers, and 23 percent when operated with ABS-equipped trailers, both statistically significant.
An initial analysis of the state data for all levels of crash severity (property damage only, or resulting in an injury or
However, the age differences between the ABS and non-ABS tractors were found to have biased the results because the non-ABS tractors were at least two years older than the ABS-equipped tractors. Additional analyses of the state data were conducted on an age-restricted subset of the crash data for overlapping tractor ages at the time of the crash for both ABS tractors and non-ABS tractors. Since varying years of state data were used, the tractor age varied between three and ten years at the time of the crash depending on the state (e.g., between three to ten years for Florida, and eight to nine years for North Carolina).
The results of the age-restricted state data still showed crash reductions for the ABS tractors in each of the seven states, but the reductions were smaller than those seen from the unrestricted data set and there were few results that were statistically significant.
A similar analysis was conducted using 50-state FARS data from 1998 to 2008 with a data set of 30,275 crashes. The analysis considered tractors towing one trailer, but only the effectiveness of tractor ABS was considered since trailer model year information was not available. Comparisons were conducted similarly to those in the state data analysis, with a control group consisting of crash types in which ABS would not be considered to have an influence, and a response group in which ABS could be considered to have an influence in the crash. The response and control groups included both ABS tractors and non-ABS tractors.
The initial FARS results found that the ABS tractors in the response group had an overall two percent increase in crashes compared to non-ABS tractors, although these results were not statistically significant.
In two-vehicle rear end crashes with the truck as the striking vehicle, a 44 percent increase was seen for the ABS tractors. However, there was an eight percent reduction in other multi-vehicle crashes in which the truck was the at-fault vehicle. Since there were many more multi-vehicle crashes that are in the “other,” non-rear-end crash category, the net result was a non-significant one percent increase in overall multi-vehicle crashes for the ABS tractors. In addition, the ABS tractors were found to have a slightly higher percentage of crashes occurring on wet roadways (18 percent of crashes occurring on wet roadways) compared to the non-ABS tractors (16 percent of crashes occurring on wet roadways), which was contrary to what was seen in the analysis of the state data.
The FARS data were then segregated by roadway locality and speed, and the results showed that reductions in crashes for the ABS tractors occurred on non-high-speed roadways (both rural and non-rural), while the increases occurred on high speed roadways (mainly rural, with only a slight increase on non-rural roads).
The analysis found that the type of road locality, travel speed, and ambient lighting condition (daylight or non-daylight) were influential in the fatal crash data. The data were then weighted to account for these influences and the final estimates for tractor ABS effectiveness and confidence intervals were derived.
During the rulemaking in the 1990's to require ABS on air-braked heavy vehicles (and, concurrently, to require ABS on medium and heavy trucks and buses equipped with hydraulic brakes), the agency solicited public comments and input on how the ABS requirements would be implemented, including a definition of ABS, ABS equipment requirements for different vehicle types, and ABS road tests to set pass-fail performance criteria for tractors, trucks, and buses. An advanced notice of proposed rulemaking (ANPRM) was published on June 8, 1992
On May 1, 1998, the agency issued an interpretation letter in response to an inquiry from Navistar International (Navistar) regarding air-braked vehicles that are equipped with an all-wheel drive (AWD) system that is selectable by the driver. Under this scenario, the vehicles are normally operated in two-wheel drive mode, and the AWD mode is selectable by the driver for severe service, off-road operation. Navistar asked if the ABS on such vehicles needed to be fully operational when the vehicle is in the AWD mode. The agency's letter stated that there is no exception in FMVSS No. 121 to permit the ABS to be disabled when AWD has been selected, although the ABS operation could be modified to better suit off-road conditions, as can be found in construction, logging, or mining operations for example. The requirements in S6,
The practical effect of this agency interpretation letter is that during a stopping distance test, the vehicle must comply with the stopping distance requirements and meet the wheel lockup provisions specified in the standard, and during a stability and control test the vehicle must remain in the 12-foot-wide lane during a full brake application in at least three out of four test runs, with the ABS fully functional and, if so equipped, a front drive axle or an interaxle locking system disengaged via the driver controls. However, when either a front drive axle or interaxle locking system is engaged by the driver, additional wheel lockup could be provided to meet operational needs. An example of this is a logging truck descending a steep grade on a muddy road at very low speeds, where some wheel lockup is needed to restrict the forward motion of the vehicle by allowing a wedge of mud to build up in front of the tires. Thus, a vehicle manufacturer can activate a modified ABS algorithm based upon the driver engaging the controls for an interaxle locking system or front wheel drive system as such needs are identified by the vehicle manufacturer. To date, the provisions already contained in FMVSS No. 121 permit modified ABS operation, without the need for an ABS on-off switch.
The purpose of requiring ABS on medium and heavy vehicles, including tractors and trailers, is to improve vehicle control and stability during panic braking. During normal driving, drivers brake lightly and no wheel lockup occurs. However, when faced with an imminent crash situation, drivers may apply the brakes by making a full brake pedal application, which can result in wheel lockup at one or more wheels on a vehicle. Since locked wheels cannot provide the lateral force needed to maintain directional control or to permit the driver to steer the vehicle around an obstacle, a loss-of-control situation occurs. A jackknife can occur if the tractor's drive axle wheels are locked and the tractor rotates about its center of gravity (often until it makes contact with a trailer being towed), or if the locked wheels on the trailer cause it to swing out of its travel lane. Both a jackknifed tractor and a trailer that has swung out of its lane can crash into other vehicles, skid off the road and strike roadside objects, or rollover. ABS keeps the wheels from locking up; thus lateral control of the vehicle is retained so the vehicle stays in its lane and the driver can also execute a steering maneuver to try and avoid a crash.
The March 10, 1995 final rule on heavy vehicle ABS included an appendix that provided details on heavy vehicle braking systems, tire characteristics related to lateral force and longitudinal force generation relative to wheel lockup, and explained why braking-related wheel lockup causes loss-of-control crashes on heavy vehicles.
However, since the ABS final rule was published, the agency published a final rule on July 27, 2009, which requires shorter stopping distances for truck tractors.
In terms of on-the-road stopping distance performance of tractor-trailers, ABS may also improve the stopping distance compared to a driver's best effort on a non-ABS brake system, particularly if the vehicle is not loaded optimally or if the roadway is slippery. For example, a tractor-trailer that is half-loaded with the load placed only in the forward half of the trailer would first experience trailer wheel lockup during hard braking if there was no ABS on the tractor or trailer. In order to prevent the trailer from swinging out of the lane, the driver would need to modulate the brake pedal to alternate between a momentary trailer wheel lockup condition, and an unlocked trailer wheel condition. However, if the tractor and trailer both were equipped with ABS, then the driver could apply the brakes with a higher pressure to take advantage of the greater tire traction available on the heavier-loaded tractor drive axles, and the ABS would prevent the trailer wheels from locking up. Thus, ABS allows the driver to use the peak amount of friction available at each wheel position even though the load at each wheel may vary greatly.
Under ideal loading conditions, such as a fully loaded tractor-trailer on dry pavement, a highly skilled test driver may be able to achieve the shortest possible stopping distance without activating the ABS system by braking the vehicle so that the brake pressure is just below the threshold of wheel lockup. However, on the highways when faced with an imminent crash threat, drivers often make a full brake application, thus engaging the ABS if any wheels are prone to lockup or going into a jackknife or trailer swing on vehicles without ABS. In summary, we believe that trucks equipped with ABS have improved stopping distance compared to non-ABS trucks when lightly-loaded, and particularly on wet or slippery roads. ABS also provides the driver with an increased level of confidence that he/she can make a hard brake application in crash-threatening situations and still be able to maintain directional control of the vehicle.
The agency reviewed the crash data that were cited in the petition as the basis for requesting to either vacate FMVSS No. 121, or requiring an on-off switch or automatic disabling of the ABS on heavy vehicles at speeds greater than 55 mph. The petition stated that the agency's report on the ABS effectiveness on tractors and trailers showed no statistically significant benefits in reducing fatal truck crashes and that the best estimate of a reduction in all types of crashes by having ABS on the tractor was only three percent. The petition stated that ABS increased overall fatalities by one percent. The agency finds that the overall three percent crash reduction for the data from the seven states correctly reflects the findings in the report, with overall crash reductions ranging between two percent and six percent for each state. Considering the response group of crashes in which ABS was possibly influential in the crash, the reductions in all crash types for ABS tractors ranged between three and ten percent for the seven states, with a median value of six percent, when compared to a control group of vehicles involved in crashes in which ABS would not be likely to be influential.
However, the one percent increase in fatal crashes for ABS tractors cited in the petition is from Table 15,
The petition addresses the effectiveness study's findings on the effect of ABS in selected subgroups of crashes. The agency notes that examination of subgroups is typically an important component of the agency's evaluations. Nevertheless, when the data are limited, as in this case, the results for the various subgroups typically comprise a wide range of positive and negative results, and some of the outlying results may even achieve statistical significance. However, without additional confirmation from other sources, it is not clear if such results are meaningful. They should be considered secondary to the overall effectiveness rating.
The petition cited the subgroup of two-vehicle rear end crashes with the truck as the striking vehicle in Table 20,
However, this single data result does not convince the agency that there would be any potential safety benefit to disabling the ABS at speeds greater than 55 mph, allowing drivers to disable the ABS, or removing ABS altogether on heavy vehicles. The aggregate of all fatal crash data shows a trend of tractor ABS reducing fatal crashes. Six of the crash subgroups also reflect reductions in crashes among ABS tractors, and two subgroups show increases among ABS tractors. The petition did not address specifically how ABS could be contributing to increases in fatal rear end crashes with the tractor as the striking vehicle, other than the unsubstantiated indirect effect of motor carriers retaining less qualified drivers to drive ABS-equipped tractors.
Furthermore, the state data results in Table 27,
In summary, since ABS improves vehicle control and stability and may have improved stopping distance performance during panic braking and under other circumstances, the agency is not able to explain why the crash data show an increase in fatal rear end crashes among the ABS tractors with the truck as the striking vehicle. The state data for all types of crashes involving tractor-trailers show decreases in rear end crashes among the ABS tractors in four states while three states show an increase in rear end crashes among the ABS tractors. The answer may not be related to ABS at all. However, the crash data provided no insight into possible relationships between the data and ABS performance in rear end crashes.
The petition stated that “antilock brakes reduce rollovers by preventing truckers from steering to avoid hitting cars” and alluded that this prevention of steering control caused an increase in rear end crashes with the ABS tractors as the striking vehicle. However, the agency finds that ABS prevents wheel lockup during braking so that steering control is maintained. Therefore, because trucks without ABS would not have steering control when the wheels are locked in a panic braking situation, the agency believes that they would be more likely candidates to strike leading vehicles than tractors equipped with ABS. The agency concludes that the petition incorrectly stated that tractor-trailers equipped with ABS do not have steering control; in fact they have improved steering control compared to tractor-trailers without ABS. We note, however, that if the ABS is not maintained in proper working order, it would not provide the improved steering control as designed. That is one reason that a crash data analysis on the basis of year of vehicle manufacture contains some uncertainty regarding the effectiveness of ABS, as was noted in the report.
The petition stated that drivers operating in rural areas should disable their ABS, while drivers operating in urban areas should not. The agency does not believe that it is valid to apply the subgroup results from the data analysis in reaching conclusions about whether ABS should be disabled on roads because of their locality. ABS operates identically on either type of road. There is no technical justification included in the petition explaining how disabling the ABS would reduce crashes, other than the concept that more highly skilled drivers would be required to be hired to drive trucks. The agency believes that disabling the ABS on heavy vehicles would result in an increase in crashes, based upon the overall results of the ABS effectiveness study. The only technical justification that the agency is aware of for disabling ABS to increase braking performance is to increase wheel lockup on loose surface roads under severe, off-road conditions. We note that this has already been addressed by vehicle manufacturers without the need to completely disable the ABS.
The petition stated that the agency's study was unable to explain the 21 percent increase in single vehicle trucker fatalities observed in 1997 when ABS was mandated, and speculated that this was not directly caused by ABS itself, but due to an unintended side effect of hiring less qualified drivers since ABS reduces the cost of tire damage from lockup of the truck's wheels during panic stops. The agency has not previously analyzed this yearly increase in truck occupant fatalities, and this issue was not investigated in the agency's ABS effectiveness study. However, we have reviewed the data and reached the following conclusions. Table 10,
Furthermore, the effective date of March 1, 1997 for truck tractors to be equipped with ABS only applied to newly-manufactured tractors, which would have only made up a small percentage of the total number of tractors on the road by the end of 1997. We do not have production figures for 1997 tractors but assuming that ABS-equipped tractor production was on the order of 100,000 units manufactured between March 1, 1997 and December 31, 1997, they would have constituted less than six percent of the 1,790,000 registered combination trucks on the road in 1997 (plus an additional small unknown percentage of tractors also on the road that were already voluntarily equipped with ABS prior to March 1, 1997). There were few ABS-equipped tractors on the road in 1997 so any positive (or potentially negative) safety effects of ABS would have been minimal during the first year of the ABS mandate for tractors. Thus the agency cannot attribute any ABS effects to the unusual increase in truck occupant fatalities that occurred in 1997.
As to the premise in the petition that the presence of ABS on heavy vehicles causes less-qualified truck drivers to be retained by motor carriers, when those drivers would otherwise have had their employment terminated due to a tractor jackknife crash that could occur with a non-ABS equipped tractor, the agency has no data, nor did the petitioner provide any, to support this claim. However, we believe that it is unlikely that the presence of ABS on a tractor by itself causes less-qualified truck drivers to be hired or retained. Truck driving has many professional aspects including driver physical qualifications; commercial driver's license requirements, including an air brake endorsement to operate air-braked trucks; and the Federal regulations that govern the loading and securing of cargo, vehicle inspections and maintenance.
The petition stated that the petitioner's own calculations showed that ABS probably saved the lives of 12 percent of truckers in 1998, 16 percent in 1999, and 5 percent in 2000. Here again, the agency believes that while tractors on the road were increasingly equipped with ABS as new vehicles entered service after March 1, 1997, there were still many trucks on the road that were not ABS equipped during those years. The details of the petitioner's analysis were not included
The agency has reviewed the petition and is denying it. The agency does not plan to initiate rulemaking or other actions to consider removing ABS from heavy vehicles, to consider requiring an on-off switch for the driver to disable the ABS, or to consider requiring the automatic disabling of ABS at speeds greater than 55 mph. The petitioner has not demonstrated that a safety need exists, which would justify removing or disabling ABS on heavy vehicles, or to vacate FMVSS No. 121 or the ABS requirements contained in it. The safety-need basis of the petition included citations of the agency's study on the effectiveness of ABS on tractor-trailers, and a claim that ABS has allowed less-skilled truck drivers to operate trucks. However, citing a subgroup of FARS data where there was an increase in fatal rear-end crashes among ABS tractors on a particular type of roadway (
The petition cited a slight increase in overall fatal crashes among ABS tractors, but when those data were weighted to account for the effects of road type and lighting condition, the results indicated an overall reduction in fatal crashes. Although this result was not statistically significant, possibly due to the limited amount of available crash data, the results of the study indicated that ABS is effective in reducing all crashes, with quite possibly a similar effect on fatal crashes. Beyond these data that were cited in the petition, there was the claim that ABS allows incompetent truck drivers to operate trucks. The agency concludes that while there are variations in levels of experience of truck drivers, they all must meet the same qualifications to drive trucks. We do not believe that ABS somehow allows incompetent drivers to drive trucks.The agency notes that, since the ABS final rule was published in 1995, only one ABS functionality problem has been identified related to some trucks operating in severe, off-road conditions. This problem has been resolved by using a modified ABS algorithm to provide an additional amount of wheel lockup at very low vehicle speeds. The vehicle manufacturers can incorporate this feature as needed by switching to a modified ABS wheel slip algorithm when a front drive axle or interaxle locking system is engaged by the driver. The agency is not aware of any other functionality problems with heavy vehicle ABS that would justify disabling it. We conclude that the petition has not demonstrated that there is a safety need or other technical reason that would justify disabling the ABS at highway speeds under any circumstances.