It's a massively disruptive technology that once it works, will just exponentially increase in use. 2019 is when Telsa will be releasing the Tesla Semi Truck which has autonomous features, numerous large companies in the USA have ordered them including Wallmart which own Asda. Give it a year or so of trials and rapid improvements by Telsa thanks to the huge amount of data they will be getting and its not at all impossible that autonomous trucks will be on the road by 2020/2021, most likely for the easier trips like distribution centre to store, probably with a baby sitter on board to take over control once it gets into a town or where needed.
Volvo already have some for trial that are in use at a quarry, they are operating between quarry and a port, presently they operate with a "supervisor" but they are generally completely anonymous and could operate 24/7.
https://www.cbronline.com/news/autonomous-truck
Tesla don’t have level 3 automation, let alone level 4/5. The autonomous technology in the Semis will be a glorified autopilot seen in the Model S/3, a driver assist technology, not an autonomous technology. That’s not to say autonomous lorries are not in the future. I’m guessing by 2025 there will be numerous automated lorries being tested and working small scale on public roads on certain routes. Whether it’s Tesla is another matter. Their level 2 technology is pretty good, but it’s just that. There are lots of companies playing with level 4 systems however (including companies like Volvo).
As for autonomous lorries/trucks/machines off public roads. Many quarries have been using automated dump trucks and other vehicles for the best part of a decade. Many of them without anyone in a cab at all.
There are several things to consider when talking about autonomous systems and the future of car ownership:
The technology and systems - things like the use of cameras and lidar, but also where companies are working. Companies like Tesla are starting from the “highway” side - driving on simple, defined roads at high speeds, companies like Wayno are starting from the “town” side of things, where systems have to be far more complex to deal with all the variables. So far there’s no evidence that those systems that work well in simple “highway” environments (like Autopilot) will translate to working well in town situations.
Public requirements also factor in. As significant proportion of the population need/want more than a ridesharing system will provide. Having to remove all your belongings every trip (and store them), find specialist vehicles (for example if you want to carry bikes or other equipment) and people that just require a lot of “stuff” in their vehicles. Hailing a vehicle works fine if you’re just commuting to work, but for a family or more specialist vehicle it’s going to be such a pain to use (where do you store your child seats when you go to the shops for example), or do you hire the vehicle for the entire time you are there (extra expense). Now think about doing that every day? Why not just lease/buy?
And remember, while automation is going to come to taxi fleets first, it’ll become available in public vehicles pretty quickly after. Why not buy/lease an automated, electric vehicle and have all the benefits of the future ride hailing vehicle? You end up with the same situation as we have now, which leads to the question - why do people still own cars now, when Uber is so cheap?
The people that believe AV is going to wholesale change the car ownership model always like to compare AV ride hailing vehicle against current ICE vehicle. Sure, it looks great, but once you compare EV AV personal vehicle against EV AV ride hailing vehicle the benefits are just not as strong, especially when you start considering practicality. If money wa small that matters then more people would take public transport and drive around in superminis. They don’t though, in part because of the practicality reasons mentioned above.
(finally!)