Make your robotaxi predictions for the 8/8 reveal

So Elon says that Tesla will reveal a dedicated robotaxi vehicle on 8/8. What do you think we will see? Will it look like this concept art or something else?

I will say that while this concept drawing looks super cool, I am a bit skeptical if it is practical as a robotaxi. It looks to only have 2 seats which would be fine for 1-2 people who need a ride but would not work for more than 2 people. I feel like that would limit the robotaxis value for a lot of people. Also, it would likely need a steering wheel and pedals for regulatory reasons even if Tesla did achieve eyes-off capability.

So I think this is concept art for a hypothetical 2 seater, cheap Tesla, not a robotaxi.

Could the robotaxi look more like this concept art but smaller? It could look a bit more like say the Zoox vehicle or the Cruise Origin, more futuristic box like shape IMO and seat 5-6 people.

Or maybe the robotaxi will look more like the "model 2" concept:

Other questions:
- Will the robotaxis be available to own by individuals as a personal car or will it strictly be owned by Tesla and only used in a ride-hailing network?
- What will cost be?
- Will it have upgraded hardware? Radar? Lidar? additional compute?
- Will Elon reveal any details on how the ride-hailing network will work?

Thoughts? Let the fun speculation begin!

 

[karlos22]

I like the idea as someone else posted here, dedicated robo taxi (maybe a model 2) which comes with FSD but needs a human driver for supervision. Steering wheel is removable. Prob lease the car directly from Tesla or some sort of ride share deal where you get the car cheaper but Tesla takes a cut.

 

[cwerdna]

As pointed out, many states don’t even have a permitting process and in the ones that do, like CA, anyone can get a non-commercial permit to deploy driverless once proven reliable with safety drivers. Sometimes there is hardly any process at all, like in NV.

So no, the permits won’t be an issue.

Waymo scales carefully because they prioritize safety. Look at Cruise. One major incident = game over.

Tesla likely won’t deploy a driverless car this decade. Probably never on vision only.

Look at the amount of scrutiny that Cruise and Waymo (esp. Cruise) have gotten over a relatively small number of significant incidents.

And even within the state of CA, you had local govts pushing back. Examples below:

SF City Attorney Files Motion to Halt Roll-Out of More Self-Driving Cruise and Waymo Robotaxis

City Attorney David Chiu is ramping up the pushback against Waymo and Cruise self-driving robotaxis in SF, filing a motion with the state to reverse their approval of unlimited robotaxis on San Francisco streets.

sfist.com

CPUC suspends Waymo's expansion requests in San Mateo, LA counties

State regulators have suspended Waymo’s robotaxi requests in San Mateo and Los Angeles counties.

www.nbcbayarea.com

Waymo expansion to SF Peninsula, Los Angeles approved

SAN FRANCISCO (KRON) — Waymo’s proposed expansion to deploy driverless autonomous vehicles to the Peninsula and Los Angeles has been approved by the California Public Utilities Commissi…

www.kron4.com

I can't speak to permitting outside CA, NV and AZ. Just because a place might not have one now doesn't mean there aren't barriers and there won't be one in the future. What if every robotaxi (safety driver or not) collision that involved significant injury or death of human or animal or that resulted in significant property damage got the amount of scrutiny and had consequences as serious as we've seen for Cruise?

Can you imagine what would happen if (insert name of robotaxi company) ended up in court (civil and/or criminal) over an accident? Imagine how many personnel at the robotaxi company would be tied up and for how long? Imagine the discovery process and cost of lawyers (and their support staff) in addition to the time of the people who get tied up?

 

[JulienW]

I think a 2 seater will have very limited appeal, regardless of it being a robotaxi or not......

So if you needed a ride you would be unhappy because it didn't have at least 3 unused seats?

 

[diplomat33]

While I admire Waymo for getting its tech working years ago..you have to wonder if there is something fishy going on since they have not expanded very quickly across the US.

There is nothing fishy at all. Launching a real driverless ride-hailing service that is actually safe, convenient and reliable is very expensive and takes time if you want to do it right. There is mapping, talking to community leaders to get their support, working with law enforcement and first responders so that they are prepared, getting permits, buying and retrofitting the robotaxis, testing the autonomous driving to make sure it can drive safely, improving the software to handle any edge cases you missed, validating that safety, ensuring your robotaxi can handle any special rules that the city has, setting up remote assistance and customer service to handle issues, setting up place to store, clean, charge, maintain, repair robotaxis in their off time, etc...

Now you can try to cut corners to deploy faster. That is what Cruise did. They did not do enough validation and did not work with the community or first responders enough. And we saw the disastrous results.

You can also take the Tesla route and deploy "beta" software and require supervision. Tesla is just having all their customers be safety drivers. So Tesla could basically do Uber with safety drivers. That would be a lot quicker and easier to scale. But it is also not a true driverless robotaxi service.

 

[Ni Ph]

Maybe Elon should get his Son to draw him a design for Robotaxi too. Seemed to work well with the CyberTruck

 

[JulienW]

There is nothing fishy at all. Launching a real driverless ride-hailing service that is actually safe, convenient and reliable is very expensive and takes time if you want to do it right. There is mapping, talking to community leaders to get their support, working with law enforcement and first responders so that they are prepared, getting permits, buying and retrofitting the robotaxis, testing the autonomous driving to make sure it can drive safely, improving the software to handle any edge cases you missed, validating that safety, ensuring your robotaxi can handle any special rules that the city has, etc...

But the problem SHOULD be is Tesla is likely not going for a very limited ODD in only a few very limited geofenced areas with a small number of cars. I suspect Tesla is going to say it is shooting for 100 of thousands of robotaxies in a few years everywhere for everybody and not just for San Francisco elites type message. So a nearer limitless ODD L4 (almost L5). This makes the robotaxi Tesla is wanting to introduce a HUGE hurdle that is going to be one of the biggest technical challenges ever. And this doesn't even included all the state and local laws that will need to be addresses.

I bet as of now Tesla was not really ready to rollout a robotaxi roadmap so soon but got "Eloned" and will be scrambling to put a plain together. And of course the designers need to settle and prepare a prototype. Could be that even then Tesla will not say exactly what the final hardware suite will be since production will be >year away.

Of course Tesla may announce a nearly limitless ODD but be "forced" to start small like Waymo and add areas.

What ever happens 8/8 promises to be entertaining and news worthy.

Just a little morning rambling

 

[JulienW]

Some more rambling.

We know Tesla is working on induction charging. So it makes sense that a robotaxi would have this. BUT will Tesla make it L2 or will they make a fundamental change and offer AC Fast Charging? This will of course require a significant update to the rectifiers Tesla installs in the cars.

If the robotaxies are going to be intercity/interstate they will need AC Fast Charging induction stations installed across country. Just one more hurdle.

 

[spacecoin]

Look at the amount of scrutiny that Cruise and Waymo (esp. Cruise) have gotten over a relatively small number of significant incidents.

And even within the state of CA, you had local govts pushing back. Examples below:

SF City Attorney Files Motion to Halt Roll-Out of More Self-Driving Cruise and Waymo Robotaxis

City Attorney David Chiu is ramping up the pushback against Waymo and Cruise self-driving robotaxis in SF, filing a motion with the state to reverse their approval of unlimited robotaxis on San Francisco streets.

sfist.com

CPUC suspends Waymo's expansion requests in San Mateo, LA counties

State regulators have suspended Waymo’s robotaxi requests in San Mateo and Los Angeles counties.

www.nbcbayarea.com

Waymo expansion to SF Peninsula, Los Angeles approved

SAN FRANCISCO (KRON) — Waymo’s proposed expansion to deploy driverless autonomous vehicles to the Peninsula and Los Angeles has been approved by the California Public Utilities Commissi…

www.kron4.com

I can't speak to permitting outside CA, NV and AZ. Just because a place might not have one now doesn't mean there aren't barriers and there won't be one in the future. What if every robotaxi (safety driver or not) collision that involved significant injury or death of human or animal or that resulted in significant property damage got the amount of scrutiny and had consequences as serious as we've seen for Cruise?

Can you imagine what would happen if (insert name of robotaxi company) ended up in court (civil and/or criminal) over an accident? Imagine how many personnel at the robotaxi company would be tied up and for how long? Imagine the discovery process and cost of lawyers (and their support staff) in addition to the time of the people who get tied up?

Yes. You're making my point. It's an engineering problem. Make it safe, prove it's safe. Deploy. Improve, Educate, Repeat.

 

[DanCar]

While I admire Waymo for getting its tech working years ago..you have to wonder if there is something fishy going on since they have not expanded very quickly across the US.

The fishy part is:

1. They have to drive every street 6 times to get hi-res maps.
2. They have to program in the fine details of local ordinances which vary state by state, county by county, city by city.
3. They test, retest, and test again.
4. Deciding on safe drop off points isn't an easy task.
5. Once everything is ironed out, they have to get the cars. Very expensive. Each car cost waymore than $100K to procure and modify with its hardware.
6. As diplomat said, buying warehouses to store the cars and service them isn't a trivial task.
7. I wonder if getting Google datacenter allotments for all their needs is sometimes an issue.

 

[stopcrazypp]

So if you needed a ride you would be unhappy because it didn't have at least 3 unused seats?

Small two seaters like that have very limited passenger and cargo space. I think most people would prefer more space. For example, if it is like the robotaxies mentioned, the seats would be facing each other and the unused seats provide more legroom or space to put cargo, whereas the small 2 seater shown will be cramped even if you are the only person.

 

[rlsd]

The fishy part is:

1. They have to drive every street 6 times to get hi-res maps.
2. They have to program in the fine details of local ordinances which vary state by state, county by county, city by city.
3. They test, retest, and test again.
4. Deciding on safe drop off points isn't an easy task.
5. Once everything is ironed out, they have to get the cars. Very expensive. Each car cost waymore than $100K to procure and modify with its hardware.
6. As diplomat said, buying warehouses to store the cars and service them isn't a trivial task.
7. I wonder if getting Google datacenter allotments for all their needs is sometimes an issue.

"As diplomat said, buying warehouses to store the cars and service them isn't a trivial task."

I am willing to rent my driveway for robotaxi cars to park.

 

[JulienW]

Small two seaters like that have very limited passenger and cargo space. I think most people would prefer more space. For example, if it is like the robotaxies mentioned, the seats would be facing each other and the unused seats provide more legroom or space to put cargo, whereas the small 2 seater shown will be cramped even if you are the only person.

I agree with the "like that" example. But it is just someone's ida and no connection to Tesla. I'm sure Tesla is going for more of a "SUV" than low slung and tight space "spots car" style. As long as it is higher up/easer to enter/exit, roomy, with lots of extra luggage space and easy low/flat loading hatch it still could be a 2 seater since most taxi rides are for a single person or a couple.

 

[borland]

It will be like Johnny Cab in the movie Total Recall. Tesla using FSD (Supervised) with Optimus Robot doing the driver supervision. Additionally, Optimus robot will handle humanoid tasks for Supercharging, and customer needs.

 

[Joey D]

Tesla relies on its vision system, that’s great in a Texas or Arizona

In London I can’t remember 2 sunny days in a row since December, the vision lenses won’t cope in our bad narrow roads or cold damp weather. And the biggest obstacle ironically is other human driving on the same roads where ability and safety are so variable.

How are meat bags (aka "people") able to drive in such conditions?

 

[spacecoin]

How are meat bags (aka "people") able to drive in such conditions?

With their brain, movable eyes and movable head, ears and so on. Unfortunately the brain bit isn't yet available for machines.

 

[petercc9]

Tesla relies on its vision system, that’s great in a Texas or Arizona

In London I can’t remember 2 sunny days in a row since December, the vision lenses won’t cope in our bad narrow roads or cold damp weather. And the biggest obstacle ironically is other human driving on the same roads where ability and safety are so variable.

How often do you get deep snow or have landscapes that looks like the inside of a freezer?

You aren't that only ones with tough driving issues.

 

[petercc9]

There is nothing fishy at all. Launching a real driverless ride-hailing service that is actually safe, convenient and reliable is very expensive and takes time if you want to do it right. There is mapping, talking to community leaders to get their support, working with law enforcement and first responders so that they are prepared, getting permits, buying and retrofitting the robotaxis, testing the autonomous driving to make sure it can drive safely, improving the software to handle any edge cases you missed, validating that safety, ensuring your robotaxi can handle any special rules that the city has, setting up remote assistance and customer service to handle issues, setting up place to store, clean, charge, maintain, repair robotaxis in their off time, etc...

Now you can try to cut corners to deploy faster. That is what Cruise did. They did not do enough validation and did not work with the community or first responders enough. And we saw the disastrous results.

You can also take the Tesla route and deploy "beta" software and require supervision. Tesla is just having all their customers be safety drivers. So Tesla could basically do Uber with safety drivers. That would be a lot quicker and easier to scale. But it is also not a true driverless robotaxi service.

I don't disagree with what you are saying..but remember it has been over ~3.5 years since Waymo opened driverless cars to the general public in Chandler, AZ (and this was after YEARS of private testing)

October 8, 2020

Waymo is opening its fully driverless service to the general public in Phoenix

Five years ago this month, Steve Mahan, the retired director of the Santa Clara Valley Blind Center, who lost his sight years before, took the world's first ride in a self-driving car on public roads. This little car, called Firefly, had a top speed of just 25 mph, and used a combination of...

waymo.com

Beginning today, October 8, we’re excited to open up our fully driverless offering to Waymo One riders. Members of the public service can now take friends and family along on their rides and share their experience with the world.

Yes, they have expanded to SF and LA. But the going is slow with no signs of rapid acceleration.

 

[DanCar]

I don't disagree with what you are saying..but remember it has been over ~3.5 years since Waymo opened driverless cars to the general public in Chandler, AZ (and this was after YEARS of private testing)

October 8, 2020

Waymo is opening its fully driverless service to the general public in Phoenix

Five years ago this month, Steve Mahan, the retired director of the Santa Clara Valley Blind Center, who lost his sight years before, took the world's first ride in a self-driving car on public roads. This little car, called Firefly, had a top speed of just 25 mph, and used a combination of...

waymo.com

Yes, they have expanded to SF and LA. But the going is slow with no signs of rapid acceleration.

At least it is faster than Tesla's level 4 rollout.

 

[spacecoin]

I don't disagree with what you are saying..but remember it has been over ~3.5 years since Waymo opened driverless cars to the general public in Chandler, AZ (and this was after YEARS of private testing)

October 8, 2020

Waymo is opening its fully driverless service to the general public in Phoenix

Five years ago this month, Steve Mahan, the retired director of the Santa Clara Valley Blind Center, who lost his sight years before, took the world's first ride in a self-driving car on public roads. This little car, called Firefly, had a top speed of just 25 mph, and used a combination of...

waymo.com

Yes, they have expanded to SF and LA. But the going is slow with no signs of rapid acceleration.

And given that they use state-of-the-art in-house-developed sensors (3d perception cameras, lidar, radar) that "can see about three football fields in every direction" and they still aren't quite there yet for highway speeds should tell you all about what I believe about Tesla's robotaxi timelines.

 

[diplomat33]

Of course Tesla may announce a nearly limitless ODD but be "forced" to start small like Waymo and add areas.

If Tesla does driverless ride-hailing, I think they will need to start with geofences, just like Waymo, for the same reasons. The US is a huge country. Achieving safe driverless in the entire US, is an incredibly daunting challenge. There are just so many edge cases, safety issues, different state and local laws etc... Plus, with driverless, you need remote assistance to handle any non-safety issues. The idea that Elon seems to have that Tesla will solve FSD where they don't need remote assistance is not realistic IMO. Even super reliable driverless, will encounter edge cases and have issues from time to time, like accidentally blocking traffic, being uncertain about how to proceed or having a flat tire that will require human assistance. And imagine 1M robotaxis spread across the US. Assuming a ratio of 10 robotaxis per remote assistance operator, that would still require 100,000 remote assistance operators that need to be hired, trained, equipped with computers and communication, supervised, etc... Also, Tesla would need robotaxis depots to I don't see Tesla doing that. With geofences, it is easier to validate safety and manage fleets. Tesla could pick a city like SF, set up the geofence, overfit the data to get FSD super reliable, set up smaller remote assistance to manage the fleet.

But personally, I don't think it would make sense for Tesla to manage their own robotaxi fleet. As stated above, Tesla would need to set up remote assistance, customer support, depots to maintain the robotaxis etc... That would be a huge expense and effort that would detract from their main mission. It's why Mobileye decided not to manage their own robotaxi fleet either but instead just license the tech to others who will.

I think it makes more sense for Tesla to have Tesla customers own and manage their own "robotaxi". Tesla owners already have spaces to store, charge and maintain their cars. And Tesla owners can be safety drivers to handle any issues. So Tesla would not need remote assistance. So the more likely scenario in my mind is that Tesla creates their own ride-hailing app, like Uber. I seem to remember Elon even teasing a Tesla Network, like the Uber app. Tesla owners could be like Uber drivers and provide rides to people, and use FSD. We are seeing some Tesla Uber drivers already using FSD. And as FSD gets more reliable, the number of interventions will decrease. Eventually, Tesla owners may give rides and rarely need to intervene, except for the occasional "stall event" when FSD encounters an edge cases and does not know what to do. And Elon can always dangle the promise of FSD eventually being driverless in the future.

1. They have to drive every street 6 times to get hi-res maps.

I don't know where you are getting 6 times. But regardless, I would not call this "fishy" as there is really nothing unusual about this. It is simply building maps. It is normal to drive the same route more than once to ensure the map is correct. It is a bit like painting a wall. You can paint a wall with a single coat of paint but it common to do put on several coats of paints to ensure the best paint job. Building high res maps is certainly not the biggest limiter to scaling anyway.

Yes, they have expanded to SF and LA. But the going is slow with no signs of rapid acceleration.

You are just looking at new cities added over the long term. They are showing rapid acceleration in terms of number of rides, according to the most recent CPUC data.

"Waymo robotaxis logged 1.4 million driverless ride-hailing miles between December and February, roughly a 42% increase from the prior quarter, according to data the company reported to the California Public Utilities Commission. December was the first month Waymo reported carrying more than 100,000 passengers on driverless trips — the vast majority of which were on paid rides — and the Alphabet-owned company reported similar figures for January and February. That quarterly haul of about 316,000 passengers was up 45% from the quarter covering September through November, when Waymo carried about 218,000 riders. The latest figures represent a huge jump from this time last year, when the company’s robotaxis shuttled about 13,000 people on unpaid rides."

Source: https://www.sfchronicle.com/sf/article/waymo-service-driverless-robotaxi-19386537.php