I experienced so many moments of insight as I read this book. Among them were Sam Schwartz's discussions on
-The importance of maintaining lane and road markers and protecting roads from unanticipated sources of clutter and confusion, such as road signs defaced with graffiti that would confuse an automated vehicle image processing system;
-The criticality of rethinking land use and street real estate allocations among vehicles, bicyclists and pedestrians
-The intricate understanding of the non-verbal cues pedestrians and drivers interpret when a driver is edging toward a turn and a pedestrian is trying to cross the street in the crosswalk. Mr. Schwartz's understanding of the complexity of this and the difficulty of having it be translated and used in an artificial intelligence system was masterfully conveyed.
-The challenge of having cities and towns dependent on parking violations revenue find other sources of revenue to replace lost revenue.
There were many more discussions that wove together the technological, regulatory, economic and social challenges of transitioning to autonomous vehicles. I particularly liked his discussion of the EU's Global Data Protection Regulations and the degree to which they will inhibit the data collection that any artificial intelligence system, including the one that operates with autonomous vehicles, needs to learn and improve itself.
The one subject that might have been covered in more depth is the degree to which AVs will depend on curbside communications infrastructure that has fragmented ownership and leasehold rights, obsolete rules on who can connect to a utility pole, and the potential for municipalities to overcharge those who want to build out 5G infrastructure on municipally-owned poles. However, that is a minor omission in an otherwise outstanding book.
This is the kind of book that has been written so thoughtfully that I expect to be an exceptionally useful resource 20 years from now.