Live data from Hacker News

Beginning fully autonomous operations with the 6th-generation Waymo driver

waymo.com

131–140 of 425 posts

Re: Beginning fully autonomous operations with the 6th-generation Waymo driver

#131
post #29

Earlier quoted context omitted.

Personally as much as people like to dunk on Musk, he did build several successful companies in extremely challenging domains, and he probably listens to the world-leading domain experts in his employ. So while he might turn out to be wrong, I don't think his opininon is uninformed.

I fully agree with your first point: Musk has shown tremendous ability to manage companies to become unicorns. He's clearly skilled in this domain. However, if you think about this for 2 seconds with even a rudimentary understanding of sensor fusion, more hardware is always better (ofc with diminishing marginal value). But ~10y ago, when Tesla was in a financial pinch, Musk decided to scrap as much hardware as possib…

> humans can drive with vision only, so self-driving should be able to as well

In May 2016, Tesla Model S driver Joshua Brown died in Williston, Florida, when his vehicle on Autopilot collided with a white tractor-trailer that turned across the highway. The Autopilot system and driver failed to detect the truck's white side against a brightly lit sky, causing the car to pass underneath the trailer.

Our eyes are supported by our brain's AGI which can evaluate the input from our eyes in context. All Tesla had is a camera, and it didn't perform as well as eyes + AGI would have.

When you don't have AGI, additional sensors can provide backup. LiDAR would have saved Joshua Brown's life.

Re: Beginning fully autonomous operations with the 6th-generation Waymo driver

#132
This is a lie:

>> The 6th-generation Waymo Driver is the product of seven years of safety-proven service amassed from driving nearly 200 million fully autonomous miles across the densest cores of 10+ major cities and an expanding network of freeways. Our experience as the only company operating a fully autonomous service at this scale has reinforced a fundamental truth: demonstrably safe AI requires equally resilient inputs. This deep understanding of real-world requirements is why the Waymo Driver utilizes a custom, multi-modal sensing suite where high-resolution cameras, advanced imaging radar, and lidar work as a unified system. Using these diverse inputs, the Waymo Driver can confidently navigate the "long tail" of one-in-a-million events we regularly encounter when driving millions of miles a week, leaving nothing to the imagination of a single lens.

Waymo uses remote safety drivers that they call "fleet response agents", probably to deflect from the fact that they are, indeed, remote safety drivers.

Much like phone-a-friend, when the Waymo vehicle encounters a particular situation on the road, the autonomous driver can reach out to a human fleet response agent for additional information to contextualize its environment. The Waymo Driver does not rely solely on the inputs it receives from the fleet response agent and it is in control of the vehicle at all times. As the Waymo Driver waits for input from fleet response, and even after receiving it, the Waymo Driver continues using available information to inform its decisions. This is important because, given the dynamic conditions on the road, the environment around the car can change, which either remedies the situation or influences how the Waymo Driver should proceed. In fact, the vast majority of such situations are resolved, without assistance, by the Waymo Driver.

In the most ambiguous situations, the Waymo Driver takes the lead, initiating requests through fleet response to optimize the driving path. Fleet response can influence the Waymo Driver's path, whether indirectly through indicating lane closures, explicitly requesting the AV use a particular lane, or, in the most complex scenarios, explicitly proposing a path for the vehicle to consider. The Waymo Driver evaluates the input from fleet response and independently remains in control of driving. This collaboration enhances the rider experience by efficiently guiding them to their destinations.

From: https://waymo.com/blog/2024/05/fleet-response/

Note the language: the Waymo Driver "remains in control of driving" but a Fleet Response Agent "proposes" the path.

In other words, Waymo is not "operating a fully autonomous service", nor does it seem anything has changed now, with the "sixt-generation fully autonomous Waymo Driver". It still needs human brains to take it by the hand and help it when it gets stuck in ambiguous situations that arise despite the claim that it "can confidently navigate the "long tail" of one-in-a-million events we regularly encounter when driving millions of miles a week".

Re: Beginning fully autonomous operations with the 6th-generation Waymo driver

#133

Earlier quoted context omitted.

Probably to try to assuage people who already saw this story circulating: https://www.autoblog.com/news/waymo-uses-remote-workers-in-t...

Or perhaps those who saw this blog post by Waymo itself: Fleet response: Lending a helpful hand to Waymo’s autonomously driven vehicles Much like phone-a-friend, when the Waymo vehicle encounters a particular situation on the road, the autonomous driver can reach out to a human fleet response agent for additional information to contextualize its environment. The Waymo Driver does not rely solely on the inputs it rece…

They don’t have remote drivers. Your own link says that.

> The Waymo Driver does not rely solely on the inputs it receives from the fleet response agent and it is in control of the vehicle at all times.

> The Waymo Driver evaluates the input from fleet response and independently remains in control of driving.

Re: Beginning fully autonomous operations with the 6th-generation Waymo driver

#134
post #93

Earlier quoted context omitted.

Tesla never had lidar so they didn't abandon it. Also, Tesla started FSD in 2016. The very core of their strategy was (and is) to sell $40k car with hardware capable of running FSD. Cameras are super cheap, FSD chip is reasonably inexpensive. Lidar is not. Maybe today the cost isn't completely prohibitive (I think it still is, because you need multiple lidars) but it certainly was for the first 8 years of FSD program…

> And they didn't have sugar daddy (Google) willing to burn several billions a year for many years. Tesla's market cap is $1.3 trillion. Granted the company itself doesn't have access to all of that, but surely if they wanted to spend, say, $10 billion per year on something big like FSD, they could have. > didn't have the luxury of adding $50k to the cost of the car for the hardware A little more extreme, but: Tesla…

I'm not a fan of Tesla's approach to self driving, but

> If they simply dumped an extra $50K of material into every single one of those cars without raising the price a dime, that would be only $425 billion. That's a ridiculous sum of money, but still substantially less than [their market cap of] $1.3 trillion.

That is an apples to dishwasher comparison. Money is fungible only when it's the same kind of money on both sides. You can't compare market cap like that. (Even for a company whose market cap is seemingly divorced from reality like Tesla's)

Re: Beginning fully autonomous operations with the 6th-generation Waymo driver

#135
post #93

Earlier quoted context omitted.

Tesla never had lidar so they didn't abandon it. Also, Tesla started FSD in 2016. The very core of their strategy was (and is) to sell $40k car with hardware capable of running FSD. Cameras are super cheap, FSD chip is reasonably inexpensive. Lidar is not. Maybe today the cost isn't completely prohibitive (I think it still is, because you need multiple lidars) but it certainly was for the first 8 years of FSD program…

> And given that in Austin they just reached parity with Waymo (i.e. completely unsupervised robotaxi service), they are not doing badly. There is no unsupervised robotaxi service in Austin and there won't be, for years, if ever. Just like the way "FSD" is not fully self driving and likely never will be.

According to https://robotaxitracker.com/ there are 7 unsupervised robotaxi in Austin right now.

Re: Beginning fully autonomous operations with the 6th-generation Waymo driver

#136
post #93

Earlier quoted context omitted.

Tesla never had lidar so they didn't abandon it. Also, Tesla started FSD in 2016. The very core of their strategy was (and is) to sell $40k car with hardware capable of running FSD. Cameras are super cheap, FSD chip is reasonably inexpensive. Lidar is not. Maybe today the cost isn't completely prohibitive (I think it still is, because you need multiple lidars) but it certainly was for the first 8 years of FSD program…

> And they didn't have sugar daddy (Google) willing to burn several billions a year for many years. Tesla's market cap is $1.3 trillion. Granted the company itself doesn't have access to all of that, but surely if they wanted to spend, say, $10 billion per year on something big like FSD, they could have. > didn't have the luxury of adding $50k to the cost of the car for the hardware A little more extreme, but: Tesla…

If they had done so, their financials wouldn’t have attracted investors and they wouldn’t be worth near 1.3T

Re: Beginning fully autonomous operations with the 6th-generation Waymo driver

#137
post #71

Obviously there is a huge amount of money and effort being spent on automated driving. But I cannot help thinking that this perception technology will prove very useful for robotics in general, factory, home, in space, etc. Car dynamics are fast enough to be useful across a huge number of domains. In some sense, the visionaries in this space are not thinking big enough. I want visions of mobility with a totally diffe…

the biggest buyer of robotics is the military and i really hope waymo stays out of that

Re: Beginning fully autonomous operations with the 6th-generation Waymo driver

#138
post #93

Earlier quoted context omitted.

Tesla never had lidar so they didn't abandon it. Also, Tesla started FSD in 2016. The very core of their strategy was (and is) to sell $40k car with hardware capable of running FSD. Cameras are super cheap, FSD chip is reasonably inexpensive. Lidar is not. Maybe today the cost isn't completely prohibitive (I think it still is, because you need multiple lidars) but it certainly was for the first 8 years of FSD program…

> And they didn't have sugar daddy (Google) willing to burn several billions a year for many years. Tesla's market cap is $1.3 trillion. Granted the company itself doesn't have access to all of that, but surely if they wanted to spend, say, $10 billion per year on something big like FSD, they could have. > didn't have the luxury of adding $50k to the cost of the car for the hardware A little more extreme, but: Tesla…

You can’t trade market cap for goods and services. Tesla is not exactly rolling in cash these days.

Re: Beginning fully autonomous operations with the 6th-generation Waymo driver

#139

Earlier quoted context omitted.

According to Elon, "sensor ambiguity" is a danger to the process [1], and therefore only a single type of sensor is allowed. (Conveniently ignores that there can be ambiguity/disagreement between two instances of the same type of sensor) The fact that people still trust him on literally anything boggles my mind. [1] https://x.com/elonmusk/status/1959831831668228450

Sensor fusion allows you to resolve that ambiguity, I wonder if Elon is really as in touch with this as you would expect. No single sensor is perfect, they all have their problematic areas and a good sensor fusion scheme allows you to have your sensors reinforce each other in such a way that each operates as close as possible to their area of strength. No single sensor can ever give you that kind of resilience. Sure,…

Even if it doesn't allow you to resolve the ambiguity, knowing that there is an ambiguity is extremely valuable. Say the lidar detects a pedestrian but the camera doesn't. Which one do you believe? Well, you propagate the ambiguity and take appropriate action, i.e. slow down, change lanes, etc. Don't drive through an area where there's a decent chance that you're going to kill someone by doing it.

Re: Beginning fully autonomous operations with the 6th-generation Waymo driver

#140
post #84
post #71

Obviously there is a huge amount of money and effort being spent on automated driving. But I cannot help thinking that this perception technology will prove very useful for robotics in general, factory, home, in space, etc. Car dynamics are fast enough to be useful across a huge number of domains. In some sense, the visionaries in this space are not thinking big enough. I want visions of mobility with a totally diffe…

Google Fiber was struggling for a while because cable companies are in bed with power companies and wouldn't let them run fiber through their easement areas. In fact, even cities couldn't run their own fiber. What you envision might happen in 2100+

Google was struggling because it has the attention span of a crackle addled flea. Including leaving roads in ruins because of “shallow trenching”

https://www.cnet.com/home/internet/google-fibers-secret-weap...

https://gizmodo.com/when-google-fiber-abandons-your-city-as-...

And the HN discussion

https://news.ycombinator.com/item?id=38526745

Post reply on HN