Tesla robot taxis and human drivers compared: Here’s who is better

It is also important why we have this data at all. The NHTSA (National Highway and Traffic Safety Administration) has a general practice requiring manufacturers and operators to report certain vehicle crashes when the automation was engaged immediately before the accident, precisely in order to detect problems on the road more quickly.

Take a look at what driving in such a robotaxis looks like:

In Electrek’s review, there is one additional thing that sounds inconvenient for Tesla: a good part of the collision happened while there were human “safety monitors” in the car, people who were supposed to intervene if the system went wrong. And now, at exactly the same time, information has appeared that Tesla in Austin is starting to test driving without these controllers, which will surely increase the pressure of the public and regulators.

There is another problem with the whole story, which is that key details are often redacted in the reports, so it is difficult to reconstruct from the outside what exactly happened and who made a mistake. This does not automatically mean that robotaxi is dangerous, but it does mean that it is much more difficult for the public to believe safety claims when there is not enough clear information.

In the background is the wider “robotaxi war” in the USA. Competitor Waymo, for example, publishes its own safety reports and claims that in certain comparisons there are fewer injuries and serious crashes than human trains, although they too have their problems and recalls.

In the end, this is a story that changes from week to week. The more miles Tesla actually drives without supervision and how transparently it reports incidents, the clearer it will be to see whether this is just the “painful growth” of technology or a serious signal for the end of the whole story.