Automotive Embedded AI

Description

This course provides practical training in the complete process of designing, training, optimizing, and deploying AI models for embedded systems. In the first half, students build a foundation in embedded systems and the core principles of machine learning and deep learning, and they also gain experience accelerating inference with TensorRT. The second half focuses on application: using a rover platform, students construct a camera-based sensing-inference-control loop and develop real-time object detection and lane-keeping assist functions as part of a track-driving exercise. The course concludes with a team project, "Pick Your Challenge," in which students define and address problems encountered during track driving. By the end, students will have developed both the technical skills and the engineering judgment needed to carry out an automotive embedded AI workflow.