Some studies predict that there will be 10 million self-driving cars on the road by 2020. But that seems like a tall task as there are many hurdles self-driving cars still need to overcome.

The artificial intelligence (AI) that powers self-driving cars needs many more hours of practice before it can safely take to the roads in great numbers.

And regulations certainly will be tricky. National Highway Traffic Safety Administration (NHTSA) chief Mark Rosekind recently said the agency plans to release in July 2016 documents that will serve as the foundation for national regulations for self-driving cars. However, their framework won’t prohibit states from setting additional standards.

Brad Templeton, a leading self-driving car expert who consulted with Google’s autonomous vehicle team, joined The Robotics Trends Show to discuss these challenges and more.

We also discuss the Google vs. Tesla battle, using convolutional neural networks in self-driving cars, and what will happen when an autonomous vehicle causes its first fatal accident.

Subscribe to The Robotics Trends Show in iTunes, Google Play, SoundCloud and Stitcher so that you never miss an episode!