What is remarkable about it is the removal of man from the car. For the outside public, it is very difficult to gauge the level of progress of a robocar team. Everyone posts really cool videos of their cars driving and solving various problems. The problem is that you can create such a video at almost any level of progress if you choose what you show. Therefore, we must measure teams based on the risks they are willing to take and how many people they will allow to see all aspects of the operation.
The decision to go unmanned in the vehicle meant a big presentation was made by the team to the board where they showed the vehicle was good enough to be released this way, with members of the public and no one to grab the wheel or hit emergency. stop if problem occurs. This tells us that the team made a convincing case and the quality is good — or maybe the team is being reckless, which we’ll find out soon enough. Baidu claims 32 million km of operations to date. Baidu states that while there is remote monitoring, they have about 2-3 vehicles per remote monitor, so it’s not a 1:1 ratio.
Vehicles must pick up/drop off at designated stops, not wherever there is free pavement like human drivers do. “PuDo” is its own problem that not all teams have solved yet. (Cruz got in trouble for just PuDoing on the street without pulling over, although at night that’s common for taxis.)
The other measure of the team’s own self-assessment of how far is whether they will allow the public to see random walks. Again, it’s not that difficult to take an invited member of the press with you on a pre-planned and well-rehearsed route. By allowing members of the public to drive anywhere, anytime, you show that you are confident that this will work. Some teams require riders to sign NDAs and not make videos. The more confident groups allowed anyone to post these videos. Again, this says that the company’s own testing has told them their vehicle won’t put them in a difficult position in the videos. Baidu says that riders can take and post videos of their rides, so it will be interesting to see.
It is not enough, of course, to allow this. Tesla
To a small extent, this also means that they have convinced the regulators about it, but the truth is that the regulators are not really capable of assessing the quality of a bot. Even teams figure out exactly how to do this, but they’re the only ones with much of an idea. What they dare to do shows what their own ratings said.
In the US, Waymo has been using driverless vehicles in Arizona for several years. More recently, Cruise launched such operations at night in a limited area of downtown SF, and Waymo also began operating around the clock there, but has not fully launched unmanned service.
The ability to charge money is not a major step, although it has often been highly advertised. No one is trying to run these services as a business yet. Charging money allows them to see how the public reacts to the service when they have to pay for it and to experiment with other types of charging. For now, most services just charge similar to or slightly less than Uber