|SUBJECT:||M.S. Thesis Presentation|
|TIME:||Wednesday, December 11, 2019, 3:00 p.m.|
|PLACE:||MRDC Building, 4404|
|TITLE:||Vehicle Damage Prediction after Impact|
|COMMITTEE:||Dr. Bert Bras, Co-Chair (ME)
Dr. Michael Leamy, Co-Chair (ME)
Dr. David Torello (ME)
The focus of this thesis is on using advanced driver-assistance system (ADAS) sensors available on newer year cars and applying them in a novel way to predict post-crash damage. This is split into three steps, first is to predict crush or intrusion on the exterior line using the accelerometer signal during impact. Next, using the array of parking sensors and the adaptive cruise control (ACC) radar, the impact location on the vehicle is predicted. Once the location and crush are determined, the final step is to combine the two to predict which parts are damaged on the car. To test this algorithm, publicly available crash test data is used for crush prediction and simulated ADAS data is used for location prediction. Parking sensor and ACC radar placement shows that location detection is more accurate for front and rear impacts than side impacts. Overall, there are more accurate finite element and lumped parameter models that can be used for crash prediction; however, this approach is easier to implement and does not require vehicle parameters only available to manufacturers.