Technological approaches for the identification of the impaired driver and the temporary inhibition of his vehicle from starting by virtue of an ignition interlock system are discussed. The objectives and desired characteristics of predriving performance tests are reviewed, and a comparative analysis of competing systems tested to date is presented. General Motors second-generation development, the Critical Tracking Task (CTT), implemented as an alcohol ignition interlock, is described.
The CTT requires the operator to stabilize the output of an unstable system whose level of instability increases monotonically up to the critical point of loss of control. Quantification of the results obtained in evaluating its potential to discriminate between sober and intoxicated performances shows a great deal of promise, indicating that intoxicated failure rates of 50% for blood alcohol concentrations (BAC's) at or above 0.1% and 75% for BAC's at or above 0.14% can be attained with no sober failure rates. A statistical analysis showed significant differences between the means of the sober and drunk performance distributions but no differences between the standard deviations.
A high initial rate of learning is observed, perhaps due to the very nature of the task whereby the operator is always pushed to his limit, and the scores approach a stable asymptote after approximately 50 trials. It is anticipated that the total duration of the test would not exceed 10 seconds. The above results are being further validated by a continuing program involving a broader-based sample of the driving population and performance conditions with both BAC's and driving proficiency.