Search Interpretations

Interpretation ID: nht88-1.83

TYPE: INTERPRETATION-NHTSA

DATE: 04/01/88

FROM: MICHAEL M. FINKELSTEIN -- ASSOCIATE ADMINISTRATOR FOR RESEARCH AND DEVELOPMENT

TO: CARL KAPLAN -- EXECUTIVE VICE PRESIDENT M. S. GOLDKLANG & COMPANY, INC.

TITLE: NONE

ATTACHMT: LETTER DATED 03/07/88, FROM ERIKA Z. JONES -- EPA TO ASSOCIATE ADMINISTRATOR FOR RESEARCH AND DEVELOPMENT, RE ACCEPTABILITY OF ADVANCED BRAKE LIGHT DEVICE AS AN AFTERMARKET UNIT; LETTER DATED 11/30/81, FROM FRANK BERNDT, TO KENNETH G. MOYER; LE TTER DATED 05/02/84, FROM FRANK BERNDT TO LAWRENCE F. HENNEBERGER

TEXT: Dear Mr. Kaplan:

We enjoyed our meeting with you, Mr. Shapira and Mr. Eckstein, and the demonstration of the Advanced Brake Light Device (ABLD).

The ABLD appears to have some potential for reducing the incidence of rear-end crashes, and it appears that you are addressing the problem of false indications of braking action with some success. As noted during the discussions, the choice of the pedal release speed threshold for early activation of the braking signal does present an interesting problem in setting the trade between reducing false alarms and maximizing the number of early indications of braking. Too, there is the question of how much incremental benefit is obtainable beyond that provided by the current stoplamp system including the Center High Mounted Stoplamp.

We would be interested in seeing data which may bear on the question of the effectiveness of the ABLD. The testing you have commissioned the UMTRI staff to perform appears to be a step toward the goal of obtaining such data. If the results of the work at UMTRI staff to perform appears to be a step toward the goal of obtaining such data. If the results of the work at UMTRI confirm your representations the next step would be a serious fleet study.

There is always the possibility of NHTSA funding of a project to test a concept such as you propose, however, our research budget is now particularly constrained because of recent cuts in funding levels, and the process of developing a particular project as part of an agency accepted plan of research can be lengthy. Data obtained through privately funded research performed by competent, recognized, objective investigators is accepted for review in the evaluation of a concept by the agency.

The required magnitude of a fleet study to demonstrate the effectiveness of a concept such as the ABLD depends upon several

factors which include, for example, the pretest crash rate of the fleet to be used in the study, the amount of change or difference which is to be detected, the desired degree of assurance that an observed change or difference is not a chance event and t he desired degree of assurance that a real difference would be detected in the study. As an example, you suggest that the use of the ABLD would reduce annual rear-end crashes in the U.S. from approximately 2.6 million to 1.1 million. Assuming your data to be for the year 1984, and a figure of 1,716,768 million vehicle miles of travel for that year, this is a reduction of approximately 58% from a calculated crash rate of 1.55 to .65 crashes per million miles. If it is desired to be fairly sure that an y difference observed in the study would occur by chance only 5% of the time and that a real difference would be detected 95% of the time, a sample of 41.35 crashes is required for each of the experimental and control groups of cars. Consequently, at 1. 55 crashes per million vehicle miles, a sample of 26.73 million vehicle miles would be required. Assuming that the vehicles travel 10,000 miles per year, 2673 vehicles equipped with the ABLD, and a similar number of control vehicles without the device, would be required for a 1 year study.

If any of the values chosen for the computations are changed, of course, the result may be radically different. In this context I would like to note that it appears that your estimate of the effectiveness of the ABLD may be overly optimistic and the bas eline crash rate too high. You should note that 5400 vehicles, half with and half without CHMSL, were included in the second fleet test of the CHMSL. This number was selected on the basis of estimates of crash rate and effectiveness lower than those yo u presented. To demonstrate the effect of changes in such assumptions, the required sample size was recalculated with the assumption that the ABLD would reduce crashes by 20% rather than 58%. In addition, the stringency of the statistical criteria was r educed to allow the possibility that a difference would be observed in the study would occur by chance 20% of the time when there was no real difference and that a real difference would be detected 80% of the time. In this case 5897 vehicles equipped wi th the ABLD and the same number of control vehicles would be required for a one year study.

I am enclosing copies of several documents in response to your requests during our meetings at the NHTSA and with Pat McCann in Senator Lautenberg's office. The SAE paper will provide you with an overview of the technical history of the Center High Mount ed Stoplamp (CHMSL). The two technical reports of the fleet tests of the CHMSL will give you an idea of the sample size and exposure needed to establish the effectiveness of the device. In addition, I have enclosed copies of some pages from

a report which describe statistical considerations in defining the size of sample required in a defined project. The Regulatory Impact Analysis will give you some idea of the rational background prepared before taking regulatory action. Cost/benefi t analysis of a device to be used in addition to the current stoplamp system would be based on incremental effectiveness beyond that provided by the currently required system. The pages from the Code of Federal Regulations will give you the information about the petitioning process as well as current requirements for stoplamps.

During our meetings we briefly discussed the issue of the use of your device as an aftermarket system. We asked for an interpretation of the pertinent regulations by the Chief Counsel and a copy of her response is enclosed. As you can see from the memo randum, her opinion is that the device is illegal as either original or replacement equipment. You or any other individual or group, of course, can petition the agency to change the regulations to permit the use of the device. A more immediate problem for you, however, is that this opinion limits your options for the conduct of fleet tests to evaluate the ABLD.

One option is to equip a fleet to be operated outside the United States. In this case you would have to conform to any relevant regulations of the country in which you conducted the fleet test. A second option would be to identify a vehicle manufacture r interested in the use of the device, who could petition the agency for a temporary exemption from the regulations on the grounds that it would promote the development and field evaluation of an innovative safety device. Such an exemption would cover 2 500 vehicles a year. This approach would lengthen the probable time required to conduct a fleet test, but appears to be necessary if you desire to conduct a fleet test in this country.

I hope I have provided you with the information you need. If we can be of further help, please don't hesitate to call me or Dr. Bishop.

Sincerely,

ENCLOSURES