Instructor

Textbook: none.

Course prerequisites: no strict requirements. Helpful: CSC 114 (Intro to C++), CSC 501 (Operating Systems), CSC 548 (Parallel Systems).

Course objective: This class prepares you to understand research challenges in autonomous driving. The delivery of the material will be split between lectures and research papers, the latter being presented by students. You will be introduced to a selection of fundamental concepts from embedded, real-time systems and machine learning. In parallel, students will participate in presenting research papers on topics related to and advancing on the fundamental concepts of autonomous driving.

Learning outcomes: By the end of the course, you should be able to do the following things:

• Embedded Systems. To determine hardware and software capabilities required for computation in an embedded environment for autonomous driving; to receive sensory input, devise control algorithms, and to drive actuator outputs; to reason about functional correctness of algorithms, physicals plants and their interaction; and to consider secondary objectives such as power/energy.
• Real-Time Systems. To analyze a real-time task set for autonomous driving in terms of temporal correctness, including schedulability, resource arbitration, timing analysis, hard and soft deadline, and their implications on control systems.
• Machine Learning. To determine which machine learning techniques are suitable for different autonomous driving challenges ranging from object recognition over object tracking to route planning; and from driving assistance systems to different levels of partially to fully autonomous driving; to reason about the capabilities and limitations of these systems; to consider societal impacts.