The development of semi-autonomous driving technology and the rise of remote drivers may be putting us closer to this new reality than previously thought.
Much of the discussion around the future of transportation, aside from sustainability, has been about automation. In a world with autonomous vehicles, roads will be safer – 90% of automobile accidents are the result of human error, as the proponents point out. Autonomous vehicles also promise to make society wealthier and more productive; in North America the average city devotes one third of its land to vehicle storage. The claim is that all the parking lots, driveways and garages will become obsolete because private vehicle ownership will be uneconomical – it will be cheaper to simply order a ride from an autonomous fleet when needed.
Hard resets and beta tests
How far away are we from achieving all of this? Google has had the most ambitious plan to date. The Waymo fleet, a network of fully autonomous cars equipped with enough hardware and machine learning to outsmart a raven, has supposedly come the closest to approximating and even surpassing that of a human driver. Skeptics point out that there have been numerous fatalities resulting from Waymo vehicles colliding with human drivers.
Tesla, by contrast, has focused their efforts more on semi-autonomous technologies that keep drivers safe by reducing their cognitive workload. The Automatic Lane Keeping System (ALKS) uses sensors, cameras and artificial intelligence to keep the vehicle safely within the bounds of the lane at a safe distance from other drivers. It cannot do things like change lanes or exit a roadway. The idea is that drivers will not have to “drive” the vehicle at all times – the role of the driver will be a navigator and overseer that only has to worry about lane changes and left turns. One problem with this is that it does not solve the problem of distracted driving. If motorists are incentivised to be less vigilant, how bad will it get when things go wrong?
Remotely driven vehicles are one potential solution to this. When using a remotely driven vehicle, there is a driver to act as navigator and overseer, but they are doing so from behind a desk in front of a computer screen, without needing to be in the vehicle at all. Fetch, the latest UK based ride hailing service, is offering remotely driven ride hailing services in the Milton Keynes area.
Follow the yellow brick road
One pathway toward a future with no vehicle ownership and safe transport would be to have ride hailing services powered by fleets that are almost fully autonomous, but use remote drivers to fill the holes that the self-driving algorithms cannot. For example, it could be that the Tesla algorithms become the gold standard for safe in-lane travel and the playbook used by Fetch’s remote drivers becomes the gold standard for safely navigating every vehicle on the road in Tesla’s fleet.
To be a Tesla controller would be to fill the role of both commercial pilot and air traffic controller simultaneously. Instead of having people behind the wheel, we would have teams of people working at terminals on headsets, whose sole job is to understand exactly what the AI algorithms will and will not do, and assist accordingly. This vision of the future varies substantially from the one put forth by Google and the Waymo fleet, which has no need for manual human oversight. It will be interesting to see if either of these pan out.