Electronic Sensing of Automobile Crashes for Airbag Deployment

Although the performance of most of today's airbag crash sensor systems is excellent, it is achieved by using multiple sensors located remotely from the passenger compartment. Those systems which use single sensors in the passenger compartment sometimes fail to actuate when needed or are too slow to provide optimum occupant protection. Installed airbag system cost now suggests that designers should find ways to reduce the number of sensors and the complexity of the system without suffering degradation in performance. An electronic sensor is freed from the constraints of the differential equation of motion of an electromechanical sensor because its algorithm can, in principle, be arbitrary. It can manipulate the acceleration data in new ways, but whatever algorithm is employed must be firmly grounded in some understood physical process to guarantee the ability of the algorithm to perform properly in real world events not recreated on the test track.

An algorithm is presented which de-emphasizes the use of the simple velocity calculations used by older electronic sensors and instead, uses measures of energy dissipated during a crash event to assess whether or not the level of severity of the event is sufficient to warrant airbag deployment. Explanations of the physical and mathematical basis of the algorithm and a VLSI circuit which performs the necessary digital signal processing are given and the efforts to help assure sensing reliability are discussed.