Ford plans self-driving car for ride share fleets in 2021

By Alexandria Sage and Paul Lienert PALO ALTO/DETROIT (Reuters) – Ford Motor Co plans to offer a fully automated driverless vehicle for commercial ride-sharing in 2021, the automaker announced Tuesday, expanding its efforts in driverless cars and ride sharing

Source: Ford plans self-driving car for ride share fleets in 2021

Why is a plan for self driving cars being announced for 5 years from now when there are already self driving cars on the road today? That’s a very pressing question in light of the recent Tesla accident that caused the first fatality for autonomous cars on public roads.

As, I’ve written here before, not all of the growth items of machine learning have good results. Mistakes are a part of the process and they will inevitably continue to happen. The time between now and 2021 (when Ford will launch their fleet) will be a very interesting period for self-driving cars.

There is also a secondary consideration here when it comes to these cars. As noted in the article, the motor companies (like Ford) are teaming up with the ride-hailing apps (like Uber and Lyft). This will create a new modicum in what we consider to be “car ownership”. You will no longer actually own a car, you will merely own the rights/ability to hail a car to your location and then use the car and sent it back when you are done. That will be our mode of transportation very soon.

The part that is subtly eliminated in that discussion is the human Uber or Lyft driver that you may already be accustomed to. Who do you look to when you have an accident and there is no driver at the wheel?

The part that is subtly eliminated in that discussion is the human Uber or Lyft driver that you may already be accustomed to. They will no longer be needed under this new scheme. However, that in itself raises a liability. You will now have thousands (hundreds of thousands?) of cars with no driver on their way to their next pick up location. Who do you look to when you have an accident and there is no driver at the wheel? This creates a massive liability for the motor companies in a way which they’ve never been exposed before. Even pioneering Tesla was quick to point out that the driver in their fatality was watching a Harry Potter movie during the accident. There is always some implicit liability that can be had (created?) when there is a driver at the wheel. What do you do when that driver is no longer there and your liability lies in a series of algorithms? Car companies will go from manufacturers to owners in a very short amount of time and there is a lot of liability that comes with that fundamental shift.

So, why 5 years? Part of it is to allow for the machine learning process to mature a little bit. But another part of it might be that the legal teams need to see some case law established in order to start to quantify some of the liabilities that the car companies are creating for themselves. If they believe that autonomous cars will eliminate accidents, that’s not realistic. Road conditions are every changing and machine learning algorithms can still fail – we can easily be trading one type of accident for another. It will be an interesting time between now and 2021 and I am curious to see where this will all end up.