Are robotaxis the future? Or now in the past?
Are robotaxis the future? Or now in the past?
All the robotaxi companies claim that their safety is much better than obsolete human drivers. But are they really safer given that they seem to crash so often? This clearly needs to be investigated further.
Business Insider reports on how Tesla is requiring test drivers to test Full [sic] Self Driving on public roads - and not take over if at all possible when mistakes occur. The result is that Tesla has been endangering the public by running red lights, speeding, and getting far too close to pedestrians and bicyclists.
Jalopnik reports on how a Tesla on FSD hit and killed deer, and then didn't even stop the car. This crash was due to the Tesla relying on only a vision system instead of 3D sensors such as LiDAR or ultrasonic. If FSD cannot detect a deer it certainly won't be able to see pedestrians in similar situations.
NBC4 Washington reports that the Tesla Smart Summon system is dangerous and shoddy. And even though the self driving system has always had problems, Tesla has switched away from using much needed distance sensors.
Waymo has long claimed that their robotaxis are safe and that the more they drive, the safer they will be. But the number of crashes sure has jumped according to the data from the NHTSA.
Washington Post reports on how the National Highway Traffic Safety Administration is now investigating four Tesla crashes where "Full Self Driving" was engaged. All crashes occurred when there were visibility issues. These suggest that Tesla only using computer vision instead of LiDAR is insufficient for autonomous driving. And this is of course extremely pertinent given that Tesla's stock price is now based on robotaxis, which are in turn based on technology which Tesla does not have and is not pursuing.
The San Mateo County Transportation Authority has published their 126 page draft Automated Vehicles Strategic Plan. While it is good to see that a transportation authority is trying to understand the ramifications of the technology and develop an actual plan, it is unfortunate that they are still misguided with respect to robotaxis improving safety. The key problem is that the drivers who cause safety problems by speeding are not and will not be switching over to using robotaxis. Even if robotaxis are eventually safer than human drivers, they won't make a difference if the problem human drivers won't use them.
Andrew Hawkings at the Verge wrote a truly great article on how the Tesla "robotaxi reveal" will be a joke.
SF Chronicle reports on how a problem with autonomous vehicles is that passengers are trapped when they get harassed. Certainly expect more of this to happen.
David Zipper reports in Fast Company on how the automobile industry managed to prevent this dire needed technology from being required. Identical speed assist systems are already required in Europe on new cars, even ones manufactured in the US. The automobile companies already have to provide this!
Washington Post reports on how the National Highway Traffic Safety Administration has fined Cruise $1.5 million for withholding key information from regulators.
SF Chronicle reports that Newsom puts the interest of big corporate donors above safety. He vetoed bills that would have required human operators on large trucks. He also vetoed simple yet effective reporting requirements so that the public would actually be able to know how safe autonomous are (what are they hiding??). And to make things even worse, he signed a bill that exempts robotaxi companies from receiving tickets when their vehicles do something dangerous.
SF Chronicle reports that Cruise is gearing up for a return to the Bay Area. Note that they are not actually returning, but just "gearing up" to do so. If Cruise considers this feeble attempt newsworthy then they are indeed in deep trouble.
404 Media reports on how one of their delivery robots caused a pedestrian to be knocked down and injured. The company offered Promo Codes to make it all better. Meanwhile, they are pursuing felony charger against a couple of drunk students who temporarily snatched one of their robots.
SF Examiner reports on the new Waymo "safety" dashboard, which is basically just company propaganda. If Waymo was actually interested safety then they would pivot to providing cost effective driver assist technology. And they would instead alternative forms of transit, and of course, better road design.
Washington Post reports on California legislation that would require vehicles built after 2030 to warn drivers when they are speeding. Even though this would obviously improve safety, it is being fought against by automobile manufactures and the constitutional right to recklessly speed community. But if safety is the goal, this type of technology would make much more of a difference than robotaxis, given it would apply to all new cars.
