Waymo Claims Its Driver AI Is Safer Than a Human and Backs It Up With Test Results

You can call a driverless taxi in the Phoenix area and Waymo be there.Photo: Waymo

Not all autonomous vehicle developers are created equal. There are some firms, which we’ll not name, who have taken a more relaxed approach to safety in the name of getting their product to market, and then there’s Waymo.

Waymo is a part of Alphabet, which is the parent company of Google, and it is unequivocally the furthest along in the quest to reach SAE Level 5 autonomy. It’s done this with extensive computer simulation as well as real-world testing, and its Waymo Driver program is not only operating in several cities in the US, it’s doing so with an incredible safety record.

Waymo released a blog post on Thursday outlining its benchmarking process for collision avoidance in simulated recreations of potentially avoidable fatal crashes in Chandler, Arizona (one of Waymo’s biggest hotspots) from 2008-2017. In these simulations, the Waymo Driver AI was able to avoid all of the collisions. In contrast, the benchmarked competitor, the fictional NIEON (non-impaired, with eyes always on the conflict) human driver, could not.

Now, Waymo’s claims that it’s safer than a human driver do appear, in some cases at least, to be true based on the data shown here and its historically strong showing in California’s disengagement reports, but don’t take this as gospel because the Waymo AI is still very much in development.

The important takeaway here is that there are no self-driving cars on sale to the public today. None. Don’t believe Tesla’s “Full Self-Driving” marketing nonsense or anyone else’s. There is also no such thing as a “semi-autonomous” vehicle. Anyone who says anything otherwise is either an idiot or they’re trying to sell you something.