In this online engineering PDH course, recent developments in the technology of self-driving cars (SDCs) are described.  Popular media have made many predictions about the dramatic changes self-driving cars (SDCs) will eventually bring about in society.  Some prominent people working in the field worry that these predictions grossly overstate what can reasonably be expected of SDCs and may actually discredit the entire field when SDCs fail to meet expectations. Despite this concern, these researchers believe that SDCs will eventually prove their worth but probably not in the form and certainly not as soon as the media say. This course draws on the comments and writings of these researchers to identify over a dozen problems in SDC development that have not yet been solved. The problems include not just technical issues such as the adequacy of machine learning, software validation, hardware reliability, cybersecurity, and the lack of adequate testing, but also non-technical issues such as public fears about SDCs’ lack of safety and questions of insurance and liability. The course concludes by suggesting a future consisting of slow, incremental improvements in SDCs over many years. The radical changes in society that have been predicted in the media may never be achieved, but nevertheless SDCs may result in the saving of lives of many people who would have otherwise perished in automobile accidents. This course is based on the paper, “Self-Driving Cars: What Can We Realistically Expect?” by Mark P. Rossow, 2019.

Learning Objectives

• Four questions that a human driver and an SDC must answer to drive successfully
• The role of route maps, sensors, and software in SDC navigation
• Types of sensors
• Operational Design Domain
• SAE Standard J3016 (Levels of Automation)
• The handoff problem, disengagements, fallback, and teleoperation
• Concerns related to the widespread use of SDCs
• The future of SDCs.