Data Science Colloquium Series
2 pm to 3 pm, NKN Hall
Artificial Intelligence in Automotive Product Development
Ford Motor Private Limited.
Increased computational speeds and the ability to integrate and handle large data sets has resulted in a renewed thrust in building applications based on Artificial Intelligence (AI) (Venkatasubramanian et al., 2018). The recent promise of AI and related machine learning techniques in solving problems in domains such as game playing, retail consumer experience, autonomous cars, etc. has been instrumental in motivating the engineering community in using AI for solving problems in automotive Product Development (PD) (Winkler, et al., 2019). PD phases including research, design, integration, and testing, all of which, leading up to a vehicle launch are expensive and time consuming. Virtual verification techniques based on physics-based mathematical models have been successful in reducing our reliance on make-test-break “Edisonian” tests that are prevalent in the PD phases. AI techniques could potentially be used to augment and enhance these physics-based modeling approaches. This presentation will highlight three examples related to powertrain, emission and chassis components, where machine learning techniques augmented by domain knowledge has been used to reduce engineering time.
Santhoji Katare graduated with a Ph.D. in Chemical Engineering from Purdue in 2003. After a year at Dow Chemical Company, Santhoji joined Ford Motor Company in 2004. At Ford, Santhoji has been involved in solving problems using data analytics in research, engineering, vehicle launch and quality organizations. Santhoji has been a recipient of Henry Ford Technology Award which is the highest technical award given to Ford employees globally once a year. Santhoji has published 25+ papers in international journals and has 2 patents to his credit.