NHTSA Investigating Waymo for Bad Driving

by

Waymo is under investigation. The National Highway Traffic Safety Administration (NHTSA) said its preliminary evaluation into an estimated 444 Waymo vehicles follows 22 reports of 22 incidents including 17 collisions.

The agency said in some of those cases the automated driving systems appeared to disobey traffic safety control devices and some crashes occurred shortly after the automated driving systems exhibited unexpected behavior near traffic safety control devices.

This is the latest in a series of investigations opened by NHTSA into performance of self-driving vehicles after initiated probes into General Motors Cruise (GM.N), opens new tab and Amazon.com’s Zoox < AMZN.O>.

In February, Waymo recalled 444 self-driving vehicles after two minor collisions in quick succession in Arizona, saying a software error could result in automated vehicles inaccurately predicting the movement of a towed vehicle.

Tesla FSD Status and Waymo Safety

When comparing Tesla FSD and Waymo, miles driven to accidents are not the same as critical disengagements. If there was a human driver in the above Waymo wrong way incident that would have been a critical disengagement. However, it looks like the Waymo was lucky and an actual accident was avoided despite unsafe automated driving.

Waymo reports about 17000 miles between critical disengagements. Waymo has no human safety driver in the car so how could they have a disengagement ? Two ways. Either the system recognizes a flaw and pulls over or stops OR a remote human operator takes over.

In early 2023, Tesla reported FSD Beta engaged experienced an airbag-deployed crash about every 3.2 M miles, which is ~5x safer than the most recently available US average of 0.6M miles/police-reported crash

Teslafsdtracker.com has public crowdsourced data on the status of Tesla FSD.

98% of drives with Tesla FSD 12.3.6 have no critical disengagements versus 60% or less two years ago. 70% with no disengagements at all vs 20% two years ago.

Tesla reports its performance for Autopilot crashes and safety.

In the 4th quarter, Tesla recorded one crash for every 5.39 million miles driven in which drivers were using Autopilot technology. For drivers who were not using Autopilot technology, they recorded one crash for every 1.00 million miles driven. By comparison, the most recent data available from NHTSA and FHWA (from 2022) shows that in the United States there was an automobile crash approximately every 670,000 miles.

4 thoughts on “NHTSA Investigating Waymo for Bad Driving”

  1. US Dollar 2,000 in a Single Online Day Due to its position, the United States offers a plethora bx02 of opportunities for those seeking employment. With so many options accessible, it might be difficult to know where to start. You may choose the ideal online housekeeping strategy with the vs-70 help of this post.

    Begin here>>>>>>>>>>>>>> https://richesdays21.blogspot.com/

  2. US Dollar 2,000 in a Single Online Day Due to its position, the United States nv-02 offers a plethora of opportunities for those seeking employment. With so many options accessible, it might be difficult to know where to start. You may choose the ideal online housekeeping strategy with the vs-70 help of this post.

    Begin here>>>>>>>>>>>>>> https://buildrichness04.blogspot.com/

  3. Where to taxis normally operate? In urban areas, mostly dense cities, where private car ownership is more of a burden because of parking and other issues. So this is where Musk’s dream of Robotaxis will operate too, perhaps even more than normal taxis, given the limited range possible/allowed by a driverless car (how WILL these driverless cars recharge? They can’t plug themselves in and they could be quite a distance from a charging station in cities whenever the random customer uses up a charge to the point where some software decides it’s time to go for a charge).
    So, the percent of city drives without a DE is <70% and the city miles driven before a DE is only 15, which is basically a couple of hails here in NYC. And that's only on version 12.3.6 which has been tested just for 16 days, which is completely insufficient, even to test for all kinds of weather (which effects range too). Supposedly it's 158 miles between CRITICAL DE but it's unclear what the difference is between a "normal" DE and a Critical DE at just 15 miles – neither of which can happen in a true driverless Tesla like a real robotaxi (Tesla is already under DOJ investigation for making misleading claims about autonomous driving, even before robotaxi claims were made).
    Let's examine what it means to disengage: a human driver has to sit with his hands constantly at the ready (how many women testers are there? I've never seen any; it may matter in a panic situation, or even in a volunteer driver situation; it's well proven that women don't/won't drive in as risky situations as men, and their lower accident rate is the result). My arms get tired if I don't have an arm rest now on a long drive (I'm 65 so that is a factor). It would be impossible for me to hover my hands in the air for more than a few miles of jerky turning city driving. I think the reaction times and even the concept of DE has to be challenged. If drivers are placing their hands on the wheel at near-critical periods, instinctively ready to take over, even if they don't perform a disengagement, can it really be said they are really letting the car drive autonomously? I know even with simple cruise control, I'm much more likely to have my foot near the brake or accelerator when I'm passing a car or truck on the highway than while cruising on a highway with no other cars or trucks nearby. Fatigue, anxiety, instinctive distrust of a possibly arbitrary (to humans) autonomous driver system, and simply bad stats where it counts most, will make robotaxis a distant dream, maybe years off, if ever.
    The "if ever" part is also due to liability issues, which have yet to even be addressed for a truly driverless car. Has the NHTSA, state or federal regulators or insurance companies ever weighed in on what it would take to accept a truly driverless car on the highway? Or city? I think they would have the same skepticism and probably more, of the scant statistics I have.
    I think Tesla is going to dive after the August robotaxi presentation, unless it's already down a lot before that. A prediction was made by a Tesla short-seller on CNBC recently that Tesla is going to 50. Tesla is the second most shorted stock. I am still long on it, because it's doing a lot of things that might yet pay off, but only with money I can afford to lose.

Comments are closed.