Waymo

[Tam]

Good point. I kind of assumed @Tam was referring to adaptive cruise control since adaptive cruise control could maybe avoid the collision as he suggests. Actual dumb cruise control would certainly hit the towed truck since it is only programmed to keep a fixed speed and cannot avoid collisions. Maybe @Tam meant to say Automatic Emergency Braking, not dumb cruise control as AEB is specifically designed to avoid or mitigate collisions?

I would also add that some adaptive cruise control would also not register the towed truck (old radar ignores stopped objects) and thus would also hit the towed truck.

Early Tesla TACC was really bad at braking for an empty trailer bed or just stationary car waiting at an intersection. However, by 2018-2019 or so, it has been pretty with those scenarios.

It can still sometimes collide in those scenarios but much less.

 

[Dewg]

Early Tesla TACC was really bad at braking for an empty trailer bed or just stationary car waiting at an intersection. However, by 2018-2019 or so, it has been pretty with those scenarios.

It can still sometimes collide in those scenarios but much less.

This is due to radar vs vision. As Diplo mentioned, radar was bad at detecting stationary objects.

 

[enemji]

In the reddit AMA, Waymo confirms that they do detect road debris and potholes and will avoid as appropriate:

Interesting thing happened yesterday with my Model3. I was in the city and a black plastic shopping bag came floating in front of the car and the car slowed down to 20mph, and as the bag had no impact it then proceeded to pickup speed and go along

 

[AlanSubie4Life]

Interesting thing happened yesterday with my Model3. I was in the city and a black plastic shopping bag came floating in front of the car and the car slowed down to 20mph, and as the bag had no impact it then proceeded to pickup speed and go along

Yeah if the occupancy network detects things it occasionally tries to avoid them. I’ve seen it once or twice and I think some mysterious unclassified object shows in the visualization though I think I have never recorded it. Presumably the YouTube kids have.

 

[cwerdna]

Not sure if this is new news:

Waymo's application to expand California robotaxi operations paused by regulators | TechCrunch

Waymo's application to expand its robotaxi service in Los Angeles and San Mateo counties has been suspended for by California regulators.

techcrunch.com

Waymo’s robotaxi expansion plans in California put on hold by regulators

The hold expires in June.

www.theverge.com

 

[diplomat33]

Not sure if this is new news:

Waymo's application to expand California robotaxi operations paused by regulators | TechCrunch

Waymo's application to expand its robotaxi service in Los Angeles and San Mateo counties has been suspended for by California regulators.

techcrunch.com

Waymo’s robotaxi expansion plans in California put on hold by regulators

The hold expires in June.

www.theverge.com

It is new. Pretty disappointing if you ask me.

 

[diplomat33]

Anecdotal evidence of Waymo usage at Phoenix airport terminal.

Testing Waymo at the airport terminals tonight--in the two brief loops I made, there were like 5 or 6 cars picking up and dropping off people. Saw someone loading luggage in the back too. So cool to see real people using it for real travel purposes!!

JJ also notes that Waymo takes a back road when leaving the airport terminal right now but he assumes that will change once Waymo opens up highway driving:

Getting to the terminals is great, but leaving is a little weird and involves a weird 10 minute saunter around deserted back roads. Probably getting fixed when the freeway access rolls out (soon?™)

https://twitter.com/x/status/1760542520583532809

 

[diplomat33]

Stellantis, the automaker that owns 14 brands including Chrysler, Jeep and Ram, and autonomous vehicle technology company Waymo are not only still working together, the companies are deepening the partnership, CEO Carlos Tavares told TechCrunch in a recent interview.

This “deepened” partnership will focus on commercial self-driving Ram delivery vans, a target that was first announced in 2020 and promptly faded from public view. Discussions on this “improved” deal have focused, in part, on a crux around driverless delivery: how does the package get from the vehicle to the customer?

“When you reach the destination, how do you take the parcel out of the van?” Tavares said in a wide-ranging interview. “This has been a point of discussion that doesn’t seem easy to solve and we are now upgrading our collaboration deal with them to take that into consideration.”

https://twitter.com/x/status/1761421985215635758

 

[jabloomf1230]

https://twitter.com/x/status/1761090177135718547

 

[Daniel in SD]

https://twitter.com/x/status/1761090177135718547

Loading…

videos.files.wordpress.com

LOL.
This is good evidence that Waymo isn't using remote assistance all the time.

 

[diplomat33]

https://twitter.com/x/status/1761090177135718547

Loading…

videos.files.wordpress.com

LOL.
This is good evidence that Waymo isn't using remote assistance all the time.

For context, that incident was a year ago. As the graph shows, the number of first responders incidents have gone down to zero last month.

 

[diplomat33]

Podcast with Waymo Director of Product Management:

https://twitter.com/x/status/1762168149678432311

Here is one tidbit that I thought was interesting. Yellow lights don't always last the amount of time recommended by NHTSA. Some yellow lights are longer and some yellow lights are shorter. So Waymo uses its perception to measure how long yellow lights last and encodes that information in the HD maps so that Waymo vehicles can better predict when the yellow light will turn red, in order to avoid running a red light:

There's a feature in particular that I really love, which is that as Waymo vehicles drive around, we are measuring and recording the duration of the yellow lights in intersections that we see. Doing this isn't something that's required. If you dig into this at all, you'll see that there's a model that's published. I think it's published by NHTSA, the National Highway Transportation Safety Agency, which provides guidance to traffic engineers who are programming these lights real life, and it's based on speed, maybe the geometry of the intersection, etc.

What we found is that the model isn't always followed. For one reason or another, sometimes yellow light durations are longer, or shorter than the model suggests that should be. We always have the model that we can use, but until we have a good understanding of the real value. But once we do have an understanding of what the real duration is, that can help us make the difference between deciding to go, or deciding to stop when the light turns yellow. That's an input that we get from perception, that we log, and then we encode into our maps, and then our planning system can reason about it.

So, that when we are driving towards an intersection and the light turns yellow, we can make a much more accurate prediction of when the light is going to turn red, and therefore, whether we should stop sooner, if the duration is going to be short, so that we don't have a red light violation, or that we can continue through and have a smoother ride if it's going to be up longer. I think that's just a great example of how different parts of our system work together to produce a solution that's ultimately better for our users and for the safety of our system.

Full transcript: https://softwareengineeringdaily.com/wp-content/uploads/2024/02/SED1649-David-Margines.txt

 

[enemji]

So Waymo uses its perception to measure how long yellow lights last and encodes that information in the HD maps so that Waymo vehicles can better predict when the yellow light will turn red, in order to avoid running a red light

Might be better to get that information from the city along with HD maps. It is absurdly comedy listening to so called intelligent people showcasing their poor knowledge

 

[diplomat33]

Might be better to get that information from the city along with HD maps. It is absurdly comedy listening to so called intelligent people showcasing their poor knowledge

I am not sure who you are referring to by "intelligent people showcasing their poor knowledge". If the Waymo vehicle can get that info directly from its own sensors and store that info for later use, why not? It would seem to me better for Waymo to get the info directly, since they have that capability, rather than rely on city officials to provide the info which could be wrong.

 

[enemji]

I am not sure who you are referring to by "intelligent people showcasing their poor knowledge". If the Waymo vehicle can get that info directly from its own sensors and store that info for later use, why not? It would seem to me better for Waymo to get the info directly, since they have that capability, rather than rely on city officials to provide the info which could be wrong.

The traffic lights do not have a mind of their own like humans. They are all preprogrammed by the department of transportation. Sure ASCT is also being rolled out but most major downtowns have a very reliable historical dataset and their preprogrammed traffic lights work great. Just get that and include it in the HD maps info.

 

[diplomat33]

The traffic lights do not have a mind of their own like humans. They are all preprogrammed by the department of transportation. Sure ASCT is also being rolled out but most major downtowns have a very reliable historical dataset and their preprogrammed traffic lights work great. Just get that and include it in the HD maps info.

I get that. But I am saying why bother getting the info from the DOT when Waymo can get the info directly with their sensors? Also, one advantage of getting the info directly from the cars is that Waymo is not dependent on the DOT for the info. What if the DOT decides to change the traffic lights or the duration of the yellow light somewhere? Waymo does not want to blow through a red light because the DOT did not give them updated info. I think it is better for the vehicles to get the info in real-time and store it. That way, the info is always correct and up to date.

 

[flutas]

What if the DOT decides to change the traffic lights or the duration of the yellow light somewhere? Waymo does not want to blow through a red light because the DOT did not give them updated info. I think it is better for the vehicles to get the info in real-time and store it. That way, the info is always correct and up to date.

I mean, can't that exact same statement be applied to Waymo collecting this data?

Scenario:

Light has been timed at a 5 second yellow by Waymo.

Light is changed to be a 2 second yellow by DOT. Waymo has no knowledge of this, and therefore it's not "always correct".

A Waymo predicts the light will change in 5 seconds, and that the car will be through the intersection in 4 so they can pass through "so that we don't have a red light violation."

Well the light is now a 2 second yellow, and the car just ran a red.

 

[diplomat33]

I mean, can't that exact same statement be applied to Waymo collecting this data?

Scenario:

Light has been timed at a 5 second yellow by Waymo.

Light is changed to be a 2 second yellow by DOT. Waymo has no knowledge of this, and therefore it's not "always correct".

A Waymo predicts the light will change in 5 seconds, and that the car will be through the intersection in 4 so they can pass through "so that we don't have a red light violation."

Well the light is now a 2 second yellow, and the car just ran a red.

But how fast do you think DOT would notify Waymo and how long would it take for Waymo for update the Waymo Driver? I doubt DOT would be quick enough. It is still better for the car to measure the time of a yellow light itself directly and act accordingly. That will be faster and more generalizable.

And the Waymo Driver would likely have driving policy to be cautious regardless of how long the yellow light lasts. It is not just going to blow through yellow lights, assuming the DOT info is correct. It would likely slow down for a yellow light as the law requires.

 

[enemji]

I get that. But I am saying why bother getting the info from the DOT when Waymo can get the info directly with their sensors?

The concern I have is that Waymo is trying to build their own perceptive logic when it is already available from the DOT.

Google has already proven that they hire a bunch of high IQ people to do nothing. They hire them just so that they do not challenge google from an independent entrepreneur perspective.

An empty 3 lane road shows up as red traffic pattern if it has one stopped car in one lane. This has been like that from the day they rolled out traffic patterns on the map.

 

[diplomat33]

The concern I have is that Waymo is trying to build their own perceptive logic when it is already available from the DOT.

I think the issue is that the info from DOT is not dependable since they are not consistent with how they follow the NHTSA model. If one traffic light has a 5 second yellow light and another traffic light has a 3 second yellow light, you can't rely on that for your autonomous vehicle. And sure, you could ask the DOT to give you the traffic light logic for all the traffic lights in your geofence and embed that in your HD map but you can't rely on the DOT to update your info for your HD map every time traffic lights change. It is better for the autonomous vehicle to have its own perception logic for traffic lights, especially since it needs it's own real-time perception anyway.

Google has already proven that they hire a bunch of high IQ people to do nothing. They hire them just so that they do not challenge google from an independent entrepreneur perspective.

An empty 3 lane road shows up as red traffic pattern if it has one stopped car in one lane. This has been like that from the day they rolled out traffic patterns on the map.

Waymo is not Google. And Waymo does not use Google maps.