The landscape for automotive control is rapidly approaching different levels of autonomous driving, which creates a number of technicalchallenges, many of which the industry has not solved yet. This talkfocuses on the systems problems posed by hardware and softwaretechniques in this context. Contributions and limitations areidentified ranging from formal methods over real-time requirements forheterogeneous hardware with a need for hybrid execution andcomposition paradigms to predictable networking, fault tolerance andsecurity, ultimately leading to non-traditional techniques of deeplearning, all of which require careful consideration duringintegration. Partial solutions to these problems are provided as aninspiration, together with open problems, many of which provide ampleopportunities during the current golden age of system research posedby autonomous driving.

bio:

Frank Mueller (mueller@cs.ncsu.edu) is a Professor in Computer Scienceand a member of multiple research centers at North Carolina StateUniversity. Previously, he held positions at Lawrence LivermoreNational Laboratory and Humboldt University Berlin, Germany. Hereceived his Ph.D. from Florida State University in 1994. He haspublished papers in the areas of parallel and distributed systems,embedded and real-time systems, compilers and quantum computing. Heis a member of ACM SIGPLAN, ACM SIGBED and an ACM Fellow as well as anIEEE Fellow. He is a recipient of an NSF Career Award, an IBM FacultyAward, a Google Research Award and two Fellowships from the HumboldtFoundation.