You want to test a robocar in a simulation of everything you can think of. Now Waymo has built an AI tool with Google Gemini to make that easier to do

CES has become the smart car show. I summarize what was interesting from several companies.

Tesla has pushed a bold and risky strategy to do it with less hardware. Has this been an effort to force the team to make the harder path work?

Empty robotaxis and long drives without interventions make some imagine Tesla FSD is ready to deploy. The math may say the reverse.

A review, with video, of the new Zoox robotaxi operating in Las Vegas. The ride is good but has some flaws. Can Zoox's risky strategy win the game against Waymo, Tesla?

I don't think Waymo failed to plan for a power outage. I suspect a combination of human craziness and some Waymo mistakes made it happen

Waymo says they failed because they ran out of humans to help cars out of a jam. Why didn't simulation tell them in advance how many humans they would need?

The biggest stories of the year, with no surprise, the Waymo and Tesla constrast playing a big role in the year's news.

There are many clues that Tesla is using lots of remote monitoring and supervision, and possibly even remote driving, for their early tests. Optimus still does that.

Waymo fixed the problem, and they should definitely obey the law. But is what they did actually dangerous? And how might we redesign the school bus system in the future?

We get to see a little more of how Waymo has structured their AI as they get to do the land rush alone in the USA.

Waymo just doubled the number of cities they operate with nobody in the vehicle. How do the various teams decide where to expand next, and will it be your town?

Tesla has for years released highly misleading crash safety data for Autopilot. Now they may finally get closer to allowing apples to apples comparisons

Waymo will now drive freeways and cover most of the San Francisco peninsula, including SJC airport

Waymo seems wiling to let outsiders look at their practices and data. Where are Tesla at the others in having this important transparency? Also, what about cats?

The cost of the physical robotaxi is about 25% of the COGS of operating one, so it makes a difference, but only when competition is fierce and margins are thin

Tesla is still predicting they will go unsupervised this year in Austin. Their crash numbers don't say they should

It was no doubt scary when a flood came upon a Waymo in Phoenix last week and passengers had to walk through it to get out. Will this happen again?

One of the crashes was with a static object with minor injuries and the Tesla towed away. And a 4th crash was not reported. This does not look good.

Robots are well behaved, and should be given the best pick-up/drop-off spots at the airport when the time comes