What could possibly disrupt the great wheeled livery disruptor Uber? Self-driving cars can certainly take a chunk of money away from them for those of us who have cars, but use Uber to take us around when we go out with friends. It won’t topple Uber like they have toppled taxis, but it will pull some share.
Self-driving cars and other livery are the focus of much debate in a number of different circles. Unions, lawmakers, citizens, and technologists all have opinions on the viability of the technology.
To be clear, I love the concept. I think the technology has a ton of promise, much like many other things that technology companies are contributing to the automotive industry. But much like other problems exposed in the auto industry, there are ways to break things.
One of the things that scares me about a self-driving car is the sense of urgency to move from one place to the next. I’m not thinking about how eager I am to get to the dentist, but more in the case of an emergency. If I need to move quickly for the sake of saving life or property, will the self-driving accommodate (either automatically or via manual override)?
Take this example from Jonathan Petit. He published research that shows how a minimal investment in off-the-shelf equipment can fool poorly designed sensors in a self-driving car to think that obstacles are present when they are not. Now consider that this could be used beyond the “Imma make you late for work!” prank to the “Imma steal yo MONEY!” crime. What used to take a coordinated effort from multiple vehicles can now be done by a guy on a street corner.
Self-driving cars must follow the rules of the road, be quite resistant to remote (and local) attacks, be able to receive software updates over the air, and not cause danger to other cars—no matter if they have drivers in them or not. This list of requirements also applies. The point is to get out there and shape the discussion productively. This technology is coming to a road near you. Spend your time enabling it to do so safely.