, Professor, Mechanical Engineering
Technology that could detect a crash before it occurred would enable designers to enhance vehicle safety in a variety of ways: seat belts could pre-tighten, air bags could predictively inflate, and vehicle crush space could be enhanced by using external airbags.
Crash tests have already demonstrated that internal air bags are significantly more effective if they are given an additional predictive time of even just 30 milliseconds. Current crash detection technology relies on accelerometers that detect crashes only after or during their occurrence. The goal of this project was to design and implement a new detection system that would use anisotropic magneto-resistive sensors and short-range sonar to provide ultra-reliable detection of any imminent, unavoidable crash and localize the position of the crash around the body of the car. An analytical formulation was developed for the variation of the magnetic field around a car as a function of position. Based on magnetic field measurements using AMR sensors, the position and velocity of any other car can be estimated and an imminent collision detected just prior to collision. The developed AMR sensor system has high refresh rates, works at small distances down to zero meters, and is very inexpensive. A variety of experimental results were conducted to demonstrate the performance of the system for both one- and two-dimensional relative motion between cars. The project also conducted simulations to show the benefits of detecting an imminent collision using the developed AMR sensors, and an occupant model was developed to analyze occupant motion inside a car during a frontal collision. Finally, analytical formulations and simulations were used to show how occupant safety could be enhanced when knowledge of an imminent collision is available.
- Project number: 2010035
- Start date: 08/2009
- Project status: Completed
- Research area: Transportation Safety and Traffic Flow
Intelligent vehicles, Safety