Self driving cars are the talk of the day with more and more articles looking at the time when we can read our newspaper (assuming there will still be a newspaper -paper- to read) as our car drives us to the office.
So far self driving cars are a matter of curiosity and debate. Technology is basically here, as proven by a number of self driving cars that are around, but it is still too expensive and a few observers doubt this is the right technology. Research groups are exploring alternatives that might result in more affordable cars, like using artificial intelligence for ambient awareness -which includes image recognition- rather than LIDAR that remains quite expensive. Most recently I saw an industrial doctoral theses, in the framework of the EIT Digital Industrial Doctoral School, with a company proposing to leverage crowdsourcing images to develop 3D models that can be used by self driving cars (the advantage being of a continuous update of the ambient a car will be driving in ).
More and more cars are now equipped with an increasing number of features that eventually will make a fully autonomous self driving car: cruise control, lane control, self parking…
As predicted by Silicon Valley Mobility we can expect a growing deployment, market adoption, of assisted driving in cars (level 3 – you can text and watch a movie but still be prepared to take over) from now on, to reach full penetration around 2035/2040. In the next decade we will start seeing level 4 cars (the human driver is no longer needed in most situations and if need arise the car will come to a stop if a situation is beyond its capability). Notably about the same time that we are expected a full penetration of level 3 on cars being sold we can also expect a level 4 full penetration. As a matter of fact most of the leve 3 cars will be software upgraded to level 4. Clearly this means that level 4 will see a faster penetration time than level 3, which is to be expected since in most cases it will be a retrofit on level 3 cars upgrading them to level 4.
Level 5 cars (fully autonomous, forget the presence of a human driver) will start to be deployed in 2035 but are expected to win the market in about 10 years. Their deployment will be much faster than the level 3 and 4. This is the point where observers are expecting a market/industry disruption with a shift from the perception of the car as a personal good to own to a car as a service that can be called upon as needed.