A rapid method for identifying and characterizing structural impacts using distributed sensors: An application for automotive pedestrian protection

This research is motivated by recent activity to improve automotive safety, especially for pedestrians. In many parts of the world today, injuries and fatalities from road accidents are a significant problem. Safety features such as seat restraints and air bags provide considerable levels of protection for car occupants; however, no such protective measures currently exist for pedestrians.; Drawing upon the success and effectiveness of occupant air bag systems, current research aims to develop similar devices for pedestrians. These active pedestrian protection systems deploy a safety feature such as an external air bag when a pedestrian is hit by a vehicle. Contact with the front bumper induces a body rotation that may result in a violent head collision. The deployable safety device provides a cushioning surface for the vulnerable pedestrian during impact. The challenge of such a system is an effective sensory unit that can rapidly and correctly discriminate pedestrian impacts from non-pedestrian ones. The fast kinematics of the automobile-pedestrian impact leaves a minimal amount of time for signal processing and computation. This research study focuses on a discrimination scheme that satisfies both the time and accuracy requirements for a proposed sensory system for pedestrian protection.; A unique methodology was developed to identify structural impacts using dominant frequency features extracted from sensory data. Contact sensors mounted on the front bumper of an automobile measure the strain response from an impact event. The dominant frequencies obtained from these sensor signals are greatly influenced by the impact object's properties and can be used to discriminate between different objects. Extensive tests were conducted to gather sensor data and validate the proposed methodology and impact discrimination algorithm. Results of the impact tests indicate that the approach is sound, and the sensory system effectively identifies "pedestrian" impacts within a short period of time.