One way of avoiding crashes or mitigating the consequences of a crash is to apply an autonomous braking system. Quantifying the benefit of such a system in terms of injury reduction is a challenge. At the same time it is a fundamental input into the vehicle development process.
This paper describes a method to estimate the effectiveness of reducing speed prior to impact. A holistic view of quantifying the benefit is presented, based on existing real life crash data and basic dynamic theories. It involves a systematic and new way of examining accident data in order to extract information concerning pre-crash situations.
One problem area when implementing collision mitigation systems is being able to achieve sufficient target discrimination. The results from the case study highlight frontal impact situations from real world accident data that have the greatest potential in terms of improving accident outcome. One of the first stages in the development of forward dectection driver support systems is a system to detect vehicles travelling in the same direction. Here, braking tends to be efficient in terms of accident mitigation; hence reducing AIS1 spinal injuries in this specific traffic incidence. The need for an increased level of object discrimination becomes obvious though, when taking all frontal impacts into account.
This method forms the basis for estimating the possible benefit of autonomous braking systems as well as the basis for evaluating different settings such as sensor performance, limitation of braking time, and object classification.