3. STAKEHOLDER INTERVIEWS-BRAKE CONDITION AND MAINTENANCE AS CRASH FACTORS

3.1 Approach

As described in Section 2, published literature was sought concerning brake condition and brake maintenance as contributing factors in CMV crashes. Since availability of this type of information in the open literature was expected to be limited, a parallel effort involving interviewing of personal data sources was undertaken. The intent of these interviews was two-fold: (1) to gather data unavailable from the literature search or database analysis and (2) to corroborate the limited information that was available. The information from the surveys was then used in the economic and safety analyses.

Potential interviewees included individuals and organizations on or known to the project team, which included the CVSA, TMC, and FMCSA. Fleet and insurance contacts were provided by CVSA as well as through prior personal knowledge. Crash investigators were previously known to members of the research team. The final list of interviewees was reviewed and approved by FMCSA. Several potential interviewees indicated they were unable to respond, but referred the project team to other individuals. Survey questions were tailored to three groups of such individuals, consisting of (1) trucking or bus fleet safety representatives, (2) insurance industry representatives, and (3) crash investigators. Furthermore, during the course of the data collection, it became apparent that some of the initially proposed sources were unable to provide adequate answers to the questions. In some cases, this was because the interviewee did not have or could not obtain the requested data and in others the data could not be released due to corporate restrictions. Substitute sources were approached in these cases.

The questions were sent via e-mail in advance to allow the respondent time to gather the data, and propose a time for the interview phone call. Due to scheduling conflicts, coordination of phone interviews within the desired time frame was sometimes difficult. Therefore, the respondents were sent follow-up e-mails and left voice messages asking them to fill in the answers and e-mail them back. This was to encourage completion of the survey questions with minimum burden on the respondents' time. Once the responses were received, it was much easier for the project team to follow up with phone calls only when required for clarification. Ultimately, nine interviews were conducted, involving a range of commercial motor carrier (both truck and bus) and insurance industry representatives. Persons who conduct crash investigations (or supervise those that do) and professional accident investigators also contributed to the data, through e-mail reply to questionnaires. In these cases, for which the gathering of no new data was required, the respondent estimated a range of percentages of crashes in which brake system malfunction or failure were factors. General information on the respondents is summarized in Table 3-1.

3.2 Stakeholder Input Regarding Brake-related CMV Crashes

The responses are presented according to topic, with a tabulated summary, followed by a discussion of each group's response.

3.2.1 Brake-related Crashes

Respondents were asked about the percentage of crashes in which brake system malfunction or failure was a factor. This question was answered by the fleets based on their data from 2004, and as estimated values based over several years for the insurance industry respondents and crash investigators. A summary of their responses to this question is shown in Table 3-2.

This range of responses was surprisingly broad, with the motor carriers attributing either zero or "less than one percent" of their DOT-reportable crashes to brakes. The very low motor carrier response likely reflects the fact that the responding fleets are safe fleets with good maintenance practices. Indeed, that they participate in CVSA and were willing to share their data is an indication of their confidence in their safety record and maintenance practices. Not that they had zero crashes, as the number of DOT-reportable crashes ranged from the low 20s to over 300, depending on the fleet. Rather, they indicated with some assurance that none of their crashes involved brake system malfunction or failure. In response to a specific question, the fleets all indicated that if brakes were a contributing factor, their investigation was sufficiently detailed that they would have identified this as such.

In comparison, the insurance industry respondents estimated that the range of DOT-reportable crashes in which brake system malfunction or failure was a cause or factor was from 5 to 10 percent. This somewhat higher range is likely due to the fact that the fleets this group of respondents insures covers a range of different maintenance and safety practices, and likely have a range of safety ratings. The project team's sense is that the range reported by the insurance industry respondents is representative of the actual number, given that they deal with a wide variety of fleets, in size and maintenance. The specifics of their customers were not shared with the project team.

In contrast to these relatively low ranges, the crash investigators indicated that from 25 up to 60 percent of the crashes they had investigated could be attributed to brake system malfunction or failure. However, these numbers were admittedly high. These individuals investigate a large number of serious injury and fatal crashes annually, and are considered brake failure analysis specialists. As such, they may be called in to look at a non-representative group of crashes in which brake failure is suspected a priori. Nevertheless, their comments have provided additional insight into the question at hand. One respondent said:

We investigate over 50 CMV serious injury/fatal collisions per year and have done so for almost 20 years. As brake failure analysis is one of my specialties, my percentages of brake defects to accident cause may be higher than most. However, based on my experience, about 80 percent of the time when I am called in, I find brake defects and improper brake equipment that was not recorded by the investigating law enforcement agency. In my opinion this is a combination of:

1. Lack of training to examine CMV brake systems
2. Lack of understanding of calculating brake force relative to CMV vehicle dynamics
3. Limited enforcement power related to FMCSR
4. Limited resources - availability of DOT/CVSA inspectors
5. Limited availability of specialized test equipment to examine damaged air brake systems (not usually available to typical law enforcement personnel

For those crash investigators and testing specialists that have brake testing equipment, such as performance-based brake testers (PBBTs), at their disposal, assuming that the post-crash vehicle is capable of being tested on a PBBT, brake problems may be found at a higher rate than those from a visual inspection. Thus, the recommendations of the NTSB to certify inspectors and have drivers trained and perform pre-trip inspections of brake systems including but not limited to stroke measurements for s-cam type air brakes, without the use of similar brake force measuring equipment, would not likely provide the same benefit as if such equipment were used in maintenance and pre-trip inspections. It also should be noted that PBBTs used in the maintenance shop, or as a required part of an annual inspection, would be of increased benefit for brake types where stroke cannot be measured, such as hydraulically actuated, electric, or air-disc brakes.

One enforcement officer, while unable to provide numerical responses to our questions, did suggest that the NTSB recommendation regarding drivers' pre-trip inspection should possibly have referenced 49 CFR 392.7, since 49 CFR 396.13 relates more to the driver duties for reviewing the last daily vehicle inspection report. This recommendation has no bearing on the benefit-cost analysis, although if driver training were recommended, it would likely be applicable to both these regulations.

3.2.2 Preventable Brake-related Crashes

Respondents were also asked about the percentage of crashes that brake system maintenance or inspection could have prevented or decreased their severity. For this question, the range of responses was narrower than the first, with the fleets' estimates increasing and the investigators' decreasing, as seen in Table 3-3.

Although the range is narrower for this question, the relative ranking of respondents' answers was the same, with the fleets being the lowest and the crash investigators being the highest. There was overlap this time between the fleets and insurance companies, showing that at least some fleets felt improved brake maintenance and/or training would be beneficial in improving braking, thus preventing or decreasing the severity of crashes, as compared with their view that few crashes were the result of brake malfunction or failure.

The fact that the crash investigators' estimate was lower for this question than the previous question dealing with malfunction or failure led to a follow-up question, to which they indicated their belief that, while inspection of brakes could help decrease the number of crashes, it would not be as effective as brake performance measurement, in which malfunctions or hidden defects would be identified.

3.2.3 Additional Comments from Motor Carriers

At the close of the phone interviews, motor carrier fleet safety representatives were also asked if they had any data concerning improvements in safety or OOS statistics that they could relate directly to a specific new or improved maintenance, inspection, or training program. Although no quantitative data were provided, one respondent indicated he noticed "an improvement" with changes in maintenance and training, and another indicated a reduction in OOS rates after consolidating safety and compliance functions for five subsidiary carriers.

3.2.4 Values to be used in Economic Analysis

Although a wide range of answers were obtained from the personnel interviews, the project team applied their engineering judgment to the responses in order to identify a range of values to be used in the economic analysis. First, for each of the three groups surveyed, the range of values was consistent within each group for the two different brake-related aspects of the surveys: (1) the percentages of crashes that could be attributed to brake system malfunction or failure and (2) the percentage of crashes for which brake maintenance or inspection could have prevented or decreased their severity. To obtain the range of values to be used in the economic analysis, the lowest estimates (from the fleets) were regarded as both coming from fleets with relatively good safety ratings and also likely to be conservative. Similarly, the highest estimates (from the crash investigators) were regarded as coming from a non-representative subset of vehicles, and likely to be overestimates. The CMV insurance industry representatives were felt to have the best overall data with no reason for bias, therefore more likely to encompass an accurate and appropriate range to be used. Given these factors, for the percentage of crashes that could have been prevented or had their severity decreased from better brake inspection, we recommend using from 5 to 10 percent for a conservative estimate and from 5 to 20 percent for a less-conservative estimate. These ranges cover the overlap range of the fleets and insurance industry representatives, and the less-conservative estimate also includes the low end from the crash investigators.