Nuro Gravity testing begins

[TahoeMichael]

Uber begins early test rides of Lucid Gravity robotaxis equipped with Nuro autonomy systems

Autonomous technology expert Nuro shared the latest milestone of a partnership with Uber and Lucid Motors to develop a Gravity Robotaxi.

electrek.co

Apparently Uber employees have begun testing Nuro equipped Gravities on public roads.

 

[Dolphins4ever]

Thanks for posting! I wonder if nuro design, compute integration is better, worse, or the same as waymo. I imagine waymo has an edge due to real world data accumulation, but this is still just one part of the training. This seems ahead of schedule, even with starting with a safety monitor in the driver’s seat?

 

[TahoeMichael]

Thanks for posting! I wonder if nuro design, compute integration is better, worse, or the same as waymo. I imagine waymo has an edge due to real world data accumulation, but this is still just one part of the training. This seems ahead of schedule, even with starting with a safety monitor in the driver’s seat?

My working assumption is that Waymo is generally better, just because they've had a lot more experience with it and are now on their 6th generation of their vehicle platform. Nuro probably doesn’t need six generations to catch up, because they can learn from looking at what Waymo (and others) have done. But accumulated experience probably still matters.

 

[borski]

My working assumption is that Waymo is generally better, just because they've had a lot more experience with it and are now on their 6th generation of their vehicle platform. Nuro probably doesn’t need six generations to catch up, because they can learn from looking at what Waymo (and others) have done. But accumulated experience probably still matters.

https://ir.lucidmotors.com/news-releases/news-release-details/lucid-receive-new-investments-pif-and-uber-uber-and-lucid-expand

 

[GoDestruction]

 

[BACamster]

Spotted it on the road in Santa Clara, CA

 

[TahoeMichael]

https://ir.lucidmotors.com/news-releases/news-release-details/lucid-receive-new-investments-pif-and-uber-uber-and-lucid-expand

It's great that they're getting a substantial investment. I think it's clear that autonomous driving requires that. But did I miss something, or does that bear on how mature their technology is versus Waymo?

 

[gaskiller]

It's great that they're getting a substantial investment. I think it's clear that autonomous driving requires that. But did I miss something, or does that bear on how mature their technology is versus Waymo?

You also have look at cost to build. Waymo’s Jags are pretty expensive and need proprietary body panels. Gravity has been designed ground up for this purpose, same goes for the midsize.

If you build the vehicle yourself, it has massive advantages vs relying on another manufacturer.

 

[DeaneG]

You also have look at cost to build. Waymo’s Jags are pretty expensive and need proprietary body panels.

Waymo is using Zeeker vans now.

 

[borski]

It's great that they're getting a substantial investment. I think it's clear that autonomous driving requires that. But did I miss something, or does that bear on how mature their technology is versus Waymo?

I shouldn’t have replied to your comment; I meant to just post it on the thread.

 

[jered]

Thanks for posting! I wonder if nuro design, compute integration is better, worse, or the same as waymo. I imagine waymo has an edge due to real world data accumulation, but this is still just one part of the training. This seems ahead of schedule, even with starting with a safety monitor in the driver’s seat?

Nobody in the industry is anywhere remotely near what Waymo can do. That doesn't mean they'll sweep the market, but they have a huge head start on technology iteration and data gathering. Zoox is the only other company with fully driverless vehicles and they are much more limited today.

Can Nuro catch up? I'm not the one to answer as I don't follow this market closely, but it'll be a few years at least. Cruise was probably the closest follower to Waymo and they exited the market entirely -- partially due to a high profile accident, but largely because it was well understood the were far from Waymo's capabilities and did not have a clear path to catch up.

Why the Lucid partnership? Good question. I'd speculate that this is partially "halo effect", partially investor driven, and partially that Lucid is definitely not going to develop L3/4 in-house and Nuro's model is software licensing. I wish everyone involved the best of luck, but I think this is largely a distraction for the members of this forum -- you're not likely to get Nuro Driver on your current Gravity, nor are you likely to get a Nuro Gravity Uber anytime soon (outside of bits of the South Bay).

 

[ke2]

Waymo is using Zeeker vans now.
View attachment 38226

Not in San Francisco where Waymo's I've seen have all been either Jaguar's or Hyundai's. Zoox is also in San Francisco (usually in the south of market area from what I've been able to tell), but I've not seen Zoox (or Waymo) in the San Francisco East Bay (oakland/piedmont/berkeley/...)

 

[TahoeMichael]

Not in San Francisco where Waymo's I've seen have all been either Jaguar's or Hyundai's. Zoox is also in San Francisco (usually in the south of market area from what I've been able to tell), but I've not seen Zoox (or Waymo) in the San Francisco East Bay (oakland/piedmont/berkeley/...)

The Zeeker based Waymos (code named "Ojai") are being operated in San Francisco, but I believe they're still in the test and validation stage. So the vast majority of Waymos are the other vehicle platforms. Apparently Waymo is planning on putting the Ojais into service at some point in 2026, but I have no idea when. I've seen one driving around SF only 2 or 3 times this year so far.

 

[DeaneG]

The Zeeker based Waymos (code named "Ojai") are being operated in San Francisco,... Apparently Waymo is planning on putting the Ojais into service at some point in 2026, but I have no idea when. I've seen one driving around SF only 2 or 3 times this year so far.

We've seen several around Cupertino in the south bay.

 

[The_Don]

Nobody in the industry is anywhere remotely near what Waymo can do. That doesn't mean they'll sweep the market, but they have a huge head start on technology iteration and data gathering. Zoox is the only other company with fully driverless vehicles and they are much more limited today.

Can Nuro catch up? I'm not the one to answer as I don't follow this market closely, but it'll be a few years at least. Cruise was probably the closest follower to Waymo and they exited the market entirely -- partially due to a high profile accident, but largely because it was well understood the were far from Waymo's capabilities and did not have a clear path to catch up.

Why the Lucid partnership? Good question. I'd speculate that this is partially "halo effect", partially investor driven, and partially that Lucid is definitely not going to develop L3/4 in-house and Nuro's model is software licensing. I wish everyone involved the best of luck, but I think this is largely a distraction for the members of this forum -- you're not likely to get Nuro Driver on your current Gravity, nor are you likely to get a Nuro Gravity Uber anytime soon (outside of bits of the South Bay).

I think the current state of advancement suggests otherwise. There's a reason why a lot of folks are suggesting they'll achieve L4 by the end of this decade and it's because the technology needed to develop it has gotten really good.

I wouldn't be surprised if Nuro was in multiple cities by the end of next year. They have years of testing experience along with all that corresponding data, so they're not exactly starting from scratch.

 

[Smorgasbord]

My understanding is that Nuro is built on Nvidia's Alpamayo stack, which, thanks to Mercedes leading the way, should give them a leg up.

As for miles of data, my personal belief is that's less important with today's technology. Even Tesla uses only curated real miles of data, plus a LOT of synthetic driving miles, and reports are that only FSD disengagements where the user left a voice message explaining why are looked at by the Tesla team. They don't need more boring highway miles of data.

Alpamayo, the software for which is Open Source, includes a bunch of synthetic data. My guess is probably not enough, so companies using that need to supplement with either more synthetic or their own real data. It'll be interesting to see how mature Alpamayo really is, which would be reflected in how quickly Nuro can ramp up.

In California, the process is that they give rides to employees for a while and then ask the DMV for permission to run them, with safety drivers, for the public. Then, after doing enough of that (where enough isn't defined publicly), they ask for permission to run without the safety drivers. In Texas, it's different and easier.