Comment by jobs_throwaway
20 hours ago
Yup. And to add
> Waymo said in its blog post that its “peer-reviewed model” shows a “fully attentive human driver in this same situation would have made contact with the pedestrian at approximately 14 mph.”
It's likely that a fully-attentive human driver would have done worse. With a distracted driver (a huge portion of human drivers) it could've been catastrophic.
You're omitting the context provided by the article. This wasn't just a random scenario. Not only was this by an elementary school, but during school drop off hours, with both children and doubled parked cars in the vicinity. If somebody doesn't know what double parking is - it's when cars parallel park beside one another, implicitly on the road, making it difficult to see what's beyond them.
So you are around young children with visibility significantly impaired because of double parking. I'd love to see video of the incident because driving 17mph (27kph for metric types) in this context is reckless and not something human would typically do, because a kid popping out from behind one of those cars is not only unsurprising but completely expected.
Another reason you also slow way down in this scenario is one of those cars suddenly swinging open their door which, again, would not be particularly surprising in this sort of context.
That's my thinking as well. Taken in some abstract scenario, all those steps seems very reasonable, and in that abstract scenario we can even say it would do better than an average human would. But that is missing the overall context that this was an elementary school during drop-off hours. That's when you crawl at 3 mph expecting kids to jump behind any car, and not going at 17mph.
> But that is missing the overall context that this was an elementary school during drop-off hours. That's when you crawl at 3 mph expecting kids to jump behind any car, and not going at 17mph.
Indeed. Sure the car knows the limit, it knows it is a school zone, it can precisely track people within the reach of its sensors (but not behind blockages it can't see through).
But it is missing the human understanding of the situation. Does it know that tiny humans behave far more erratically then the big ones? Obvious to us humans, but does the car take that into account? Does it consider that in such a situation, it is likely that a kid that its sensors can't possibly detect has a high probability to suddenly dart out from behind an obstacle? Again obvious to us humans because we understand kids, but does the car know?
1 reply →
>driving 17mph (27kph for metric types) in this context is reckless and not something human would typically do
Unfortunately, a vast overestimation of human danger recognition. Or empathy, unsure
Driving is based so much off of feel so my numbers may be off, but in the scenario you are talking about 5mph seems reasonable, 10mph already seems like to much.
The want to be E but really armchair engineer in me for this context says there's far too little Engineering safety of the situation.
That school should not be on a busy roadway at all, it should also not have a child dropoff area anywhere near one but instead, ideally, a slow loop where the parents do drop off children, and then proceed forward in a safe direction away from the school in a flow.
1 reply →
> not something a human would typically do
lol I'm guessing you don't have kids. This is hilarious.
The full text says: "because driving 17mph (27kph for metric types) in this context is reckless and not something human would typically do"
While i dont have kids, i guess you dont either. Because usually kids dont drive cars, atleast i didnt when i was in elementary school.
If you drive in Sweden you will occasionally come up to a form of speed reduction strategy that may seem counterintuitive. They all add to make driving harder and feel more dangerous in order to force attention and lower speed.
One is to merge opposite directional roads into a single lane, forcing drivers to cooperate and take turn to pass it, one car at a time.
For a combined car and pedestrian road (max speed of 7km/h) near where I live, they intentionally added large obfuscating objects on the road that limited visibility and harder to navigate. This forces drivers to drive very slow, even when alone on the road, as they can't see if a car or person may be behind the next object.
In an other road they added several tight S curves in a row, where if you drive anything faster than 20km/h you will fail the turns and drive onto the artificial constructed curbs.
In other roads they put a sign in the middle of two way roads while at the same time drastically limiting the width to the curb, forcing drivers to slow down in order to center the car in the lane and squeeze through.
In each of those is that a human driver with human fear of crashing will cause drivers to pay extra attention and slow down.
In Bulgaria we have a similar speed reduction strategy but we are a bit ahead of Sweden: We use medium-radius but very deep potholes. If you lose attention for even a split second, you are forced to a full stop to change a tire. Near schools it gets more "advanced": they put parked cars on both sides of the road, and the holes positioned so you can't bypass them. For example, two tire-sized holes on both sides of the road right next to the parked cars. You have to come to a complete stop, then slowly descend into the hole with the front wheels, climb back out, and repeat the process for the rear wheels. Occasionally, even though we (technically) have sidewalks, they are covered in mud or grass or bushes, so pedestrians are forced to walk in the middle of the road. This further reduces driving speed to walking pace and increases safety in our cities. Road markings are missing almost everywhere and they put contradicting road signs so drivers are not only forced to cooperate but also to read each other minds.
Same in India! We go one better, we let people drive in the opposite lane as well!
That’s genius but one has to ask: how much does it cost to maintain these speed restricting features?
In the UK, the cost of owning a car is high yet our potholes, while frequent, are small enough to survive. Thus being more of an annoyance rather than a speed restriction.
2 replies →
It's fairly common at least in the Netherlands, Germany, and Switzerland too. In Switzerland they also place street parking spots on alternating sides on narrow streets, which also makes you more attentive and lower your speed.
I've heard that that is why roundabouts are safer than their alternatives: counterintuitively, they're safer because they're less safe, forcing the user to pay more attention as a result.
>they're safer because they're less safe
Roundabouts are safer. They're safer because they prevent everybody from speeding through the intersection. And, even in case of an accident, no head-on collisions happen in a roundabout.
5 replies →
One-lane-roundabouts are very safe. I lived in Hannover (Germany) in the 80s and 90s, they had 2 or 3 lanes in the roundabouts. There were large signs that counted the accidents (200+/year) to raise awareness and during the trade fairs (anybody remembers Cebit?) the number of accidents peaked. Today they are all a lot safer because of a lot of traffic lights.
I thought that the idea of roundabouts was that they lead to slightly more accidents than before, but they are of much lower severity than before (the 90 degree intersections they replace).
Same with driving in the winter. Anecdotally I always observe more accidents when the roads are clear.
I recently visited a friend that lives in Sweden (couple hours south of Stockholm). Something he said while I visited stuck with me:
"Sweden hates cars."
There must be a happy medium somewhere in between.
I would say it depend on where you are. City driving is generally not a great experience and its not that uncommon to see a speed bump before almost every crossing, to the point where you get surprised if there isn't one. That said, as long you don't leave the designated main roads that goes through the city areas it is not that bad. They do demand a lot of attention.
Street parking has mostly been turned into exclusive residential parking, so parking houses are often the only choice. As a result they are quite expensive, and you got to walk to the destination.
Parking and access is much better in the country side, and the highways are fairly good and similar to those found in the west Europe. It not as straight or wide as authobahn, but not as much traffic either.
It's true, Sweden isn't quite bike and pedestrian friendly enough yet, but they'll get that balance someday!
It's a runaway process of prioritizing safety over convenience -- and it's wrecking their road base just before self-driving cars would allow them to have both.
I was wondering how much convenience is worth one kid's life. This thread reminded me of some interesting terms like "value of statistical life." It appears that all those annoying low speed limits and purposeful obstructions in residential areas really do save lives.
> An evaluation of 20 mph zones in the UK demonstrated that the zones were effective both in reducing traffic speed and in reducing RTIs. In particular child pedestrian injuries were reduced by 70 per cent from 1.24 per year in each area before to 0.37 per year after the zones were introduced
https://www.rospa.com/siteassets/images/road-safety/road-saf...
The "Vision Zero" program was started in Sweden, and is becoming more widely adopted.
https://en.wikipedia.org/wiki/Vision_Zero
1 reply →
What an American framing. My convenience at the cost of your eventual safety. I guess this is why we also have toddler death machines with 5-foot grills that we call “full size” vehicles.
3 replies →
Gosh, no, the self-driving cars will be forced to drive at safe speeds in pedestrian corridors as opposed to voluntarily driving at safe speeds in pedestrian corridors. How awful.
> prioritizing safety over convenience
this sounds like exactly the right tradeoff, especially since these decisions actually increase convenience for those not in cars
1 reply →
“Just before” … this would mean all cars would be required to be self driving and that they’re forced to adhere to the set speed limits. You think this is just around the corner? In a country like Sweden with a lot of snow? Let’s talk about that this when we’re actually close to hitting 100% of self driving cars on the road.
And it’s not “runaway”, it’s exactly the right prioritisation. I’d encourage you to spend some time on Not Just Bikes and the say whether you’d like to live in a Nordic or an American neighbourhood. The Nordic style is also about convenience because car centric infrastructure makes a lot of things less accessible and convenient.
Those things all sound easy to remove in some hypothetical future where there are enough and safe enough self driving cars to have both. Makes sense to design for human driven cars for now though.
If they're actually self-driving they should be able to drive around the obstacles just as well or better than human.
Does it actually work though?
Many roads in London have parked cars on either side so only one can get through - instead of people cooperating you have people fighting, speeding as fast as they can to get through before someone else appears, or race on-coming cars to a gap in the parked cars etc. So when they should be doing 30mph, they are more likely doing 40-45. Especially with EVs you have near-instant power to quickly accelerate to get to a gap first etc.
And putting obstacles in the road so you cant see if someone is there? That sounds really dangerous and exactly the sort of thing that caused the accident in the story here.
Madness.
> Does it actually work though?
Yes. They have made steady progress over the previous decades to the point where they can now have years with zero road fatalities.
> And putting obstacles in the road so you cant see if someone is there? That sounds really dangerous and exactly the sort of thing that caused the accident in the story here.
Counterintuitive perhaps, but it's what works. Humans adjust their behaviour to the level of perceived risk, the single most important thing is to make driving feel as dangerous as it is.
4 replies →
why not just put in speedbumps if all you're trying to do is slow people down? Are you sure this was the purpose of these designs? sounds a little too freakonomics to me.
Speed bumps suck for both the driver and passangers of the car and generate road noise.
1 reply →
> It's likely that a fully-attentive human driver would have done worse.
We'd have to see video of the full scene to have a better judgement, but I wouldn't call it likely.
The car reacted quickly once it saw the child. Is that enough?
But most humans would have been aware of the big picture scenario much earlier. Are there muliple kids milling around on the sidewalk? Near a school? Is there a big truck/SUV parked there?
If that's the scenario, there is a real probability that a child might appear, so I'm going to be over-slowing way down pre-emptively even thought I haven't seen anyone, just in case.
The car only slows down after seeing someone. The car can react faster that I can after seeing someone, but as a human I can pre-react much earlier based on the big picture, which is much better.
As someone who lives on a residential street right by a primary school in the UK, the majority of drivers are going over 20mph even at the peak time when there are children everywhere.
While in theory human drivers should be situationally aware of the higher risks of children being around, the reality is that the majority will be in their own bubble of being late to drop their kid off and searching for the first free spot they can find.
the human driver would usually drive more closely to the centerline of such a residential road. If the road is clear ahead i'd drive almost over the centerline of the road having enough clearance between my path and the parked cars for any such "jumper" to be visible long enough for me to react. If there is an opposite traffic i get back strictly into my lane and slow down much more if the parked cars are close and they block sidewalk view, etc.
The autonomous cars have really got more aggressive recently as i mentioned before:
https://news.ycombinator.com/item?id=39298290
2 replies →
I vividly recall a shot within a commercial, in which a driver was shown in slow motion, chucking his coffee into the passenger foot well in order to have two hands on the wheel for an emergency. I don’t remember what was about to happen to the car or the world around it. I’m pretty sure that a collision occurred.
Have you been in a waymo? It knows when there are pedestrians around (it can often see over the top of parked cars) and it is very cautious when there are people near the road and it frequently slows down.
I have no idea what happened here but in my experience of taking waymos in SF, they are very cautious and I'd struggle to imagine them speeding through an area with lots of pedestrians milling around. The fact that it was going 17mph at the time makes me think it was already in "caution mode". Sounds like this was something of a "worst case" scenario and another meter or 2 and it would have stopped in time.
I think with humans, even if the driver is 100% paying attention and eyes were looking in exactly the right place where the child emerged at the right time, there is still reaction times - both in cognition but also physically moving the leg to press the pedal. I suspect that a waymo will out-react a human basically 100% of the time, and apply full braking force within a few 10s of milliseconds and well before a human has even begun to move their leg.
You can watch the screen and see what it can detect, and it is impressive. On a dark road at night in Santa Monica it was able to identify that there were two pedestrians at the end of the next block on the sidewalk obscured by a row of parked cars and covered by a canopy of overgrown vegetation. There is absolutely no way any human would have been able to spot them at this distance in these conditions. You really can "feel" it paying 100% attention at all times in all directions.
Your opinion of "most humans" is vastly overinflated. The median human driver would be going 5 over the speed limit, on their cell phone, and paying fuck all attention. Humans never drive as slow as 17 mph, even in the context of being directly in front of schools with visible children.
You're describing the median driver in America or India. This is not universal.
3 replies →
> But most humans would have been aware of the big picture scenario much earlier.
I wouldn't call it likely. Sure, there are definitely human drivers who are better than Waymo, but IME they're few and far between. Much more common to be distracted or careless.
When walking along a busy street facing traffic, I like to play a game of "who's using a phone?" I sometimes score in excess of 50% of drivers texting or otherwise manipulating a phone instead of actually driving.
It's amazing how much nonsense we let slide with human drivers, and then get uptight about with anything else. You see the same attitude with bicycles. Cars run stop signs and red lights all day long and nobody bats an eye, but a cyclist does it and suddenly they're a menace.
I don't think it makes sense to lump some drivers better than waymo and worse than waymo. A human brain automatically thinks of all the scenarios, where Waymo has pre-programmed ones (and some NN based ones). So it's scenarios by scenario.
Consider this scenario:
5 kids are walking on the sidewalk while you're driving past them. But suddenly a large dumpster is blocking your view of them just as you pass. You saw them before the dumpster, but not after your car and the dumpster completely blocks the view.
Does a human brain carry some worry that they suddenly decide to run and try to cross the street after the dumpster? Does Waymo carry that worry or just continue to drive at the exact same speed.
Again, it's not like every driver will think about this, but many drivers will (even the bad ones).
17 replies →
According to the article the car was traveling at 17 miles an hour before it began braking. Presumably this was in a 25 mph school zone, so it seems the Waymo was already doing exactly what you describe - slowing down preemptively.
This is close to a particular peeve I have. Occasionally I see signs on the street that say "Slow Down". I'm not talking about the electronic ones connected to radar detectors. Just metal and paint.
Here's my problem. If you follow the instructions on the sign, it still says to slow down. There's no threshold for slow enough. No matter how slow you're going, the sign says "Slow Down". So once you become ensnared in the visual cone of this sign, you'll be forced to sit stationary for all eternity.
But maybe there's a loop-hole. It doesn't say how fast you must decelerate. So if you come into the zone going fast enough, and decelerate slowly enough, you can make it past the sign with some remaining non-zero momentum.
You know, I've never been diagnosed on the spectrum, but I have some of the tendencies. lol.
18 replies →
25 mph is an upper limit, not a minimum.
A 25mph school zone? That seems fast. 15mph would be more the norm, which is in line with the 17mph the car believed itself to be traveling.
FYI, unless you are a commerical truck, a cop, or a racer, your speedometer will read slightly fast, sometimes as much as 5 to 10%. This is normal practice for cars as it limits manufacturer liability. You can check this using independant gps, ie not an in-dash unit. (Just imagine the court cases if a speedo read slower than the actual speed and you can understand why this started.)
8 replies →
In this situation, the car was already driving under the legal speed required for a school zone (25mph when children are present) [edit: some comments in the post suggest there is a 15mph sign, which is sometimes posted; to me, driving 17mph in a 15mph zone is acceptable).
I think any fair evaluation of this (once the data was available) would conclude that Waymo was taking reasonable precautions.
> was already driving under the legal speed
That's exactly part of the problem. If it is programmed to be over-cautious and go 17 in a 25 zone, that feels like it is safe. Is it?
It takes human judgment of the entire big picture to say meaningfully whether that is too slow or too fast. Taking the speed limit literally is too rigid, something a computer would do.
Need to take into account the flow of the kids (all walking in line vs. milling around going in all directions), their age (younger ones are a lot more likely to randomly run off in an unsafe direction), what are they doing (e.g. just walking, vs. maybe holding a ball that might bounce and make them run off after it), their clustering and so on.
Driving past a high school with groups of kids chatting on the sidewalk, sure 20mph is safe enough. Driving past an elementary school with a mass of kids with toys moving in different directions on the same sidewalk, 17mph is too fast.
And if I'm watching some smaller kids disappear behind a visual obstruction that makes me nervous they might pop up ahead of it on the street, I slow down to a crawl until I can clearly see that won't happen.
None of this context is encoded in the "25mph when children are present" sign, but for most humans it is quite normal context to consider.
But would be great to see video of the Waymo scene to see if any of these factors was present.
It was going 17 mph. That is rather slow.
To put it another way. If an autonomous vehicle has a reaction time of 0.3 seconds, the stopping distance from 17 mph is about the same as a fully alert human driver (1 second reaction time) driving 10.33 mph.
>It was going 17 mph. That is rather slow.
There's a case to be made that it wasn't slow enough.
5 replies →
Curiously enough Google could have access to how fast humans usually drive through that street.. if they record people's Google Maps trips, they can show the court that "Look, 80% of Google Maps users drive through here at 30 mph!".
Waymo itself has this as well. They record their drives after all which means they know the speed of the vehicles around them.
They even wrote a blog post about it:
https://waymo.com/blog/2023/07/past-the-limit-studying-how-o...
Google might even know how many drivers aren't obeying the speed limit or slow-rolling through stop signs. I wonder if they already have partnerships with law enforcement to detect areas where the traffic law is more ignored than others.
Two things:
I've read studies saying that most drivers don't brake at max effort, even to avoid a collision. This may be at least one of the reasons that Waymo predicted that an attentive human would likely have been going faster than their car at the moment of impact. I've got a good idea of my fun-car's braking performance, because I drive it hard sometimes, but after reading that I started practicing a bit with my wife's car on the school run, and... Yeah: it's got a lot more braking power than I realized. (Don't worry, I brake hard on a long straight exit ramp, when no one's behind me, a fast slow-down is perfectly safe, and the kiddo loves it.) I've now got an intuitive feel for where the ABS will kick in, and exactly what kind of stopping distance I have to work with, which makes me feel like a safer driver.
Second, going off my experience of hundreds and hundreds of ride-share rides, and maybe thirty Waymo journeys, I'd call the best 10-15% of humans better drivers than Waymo. Like, they're looking further up the road to predict which lane to be in, based on, say, that bus two blocks away. They also drive faster than Waymos do, without a perceptual decrease in safety. (I realize "perceptual" is doing some work in that sentence!) That's the type of defensive and anticipatory urban driver I try to be, so I notice when it's done well. Waymo, though, is flat-out better, in every way, than the vast majority of the ride-share drivers I see. I'm at the point where I'll choose a Waymo any time it'll go where I'm headed. This story reinforces that choice for me.
> I've read studies saying that most drivers don't brake at max effort, even to avoid a collision.
Ha! It is unbelievable how difficult it is to make someone brake hard. You'd think it's the easiest thing possible in the age of ABS - just press hard as you can.
I have a lot of experience on this, I used to teach car control both to teens and adults. One of the frequent exercises was seemingly very simple: Drive at Xmph until this spot, then brake at maximum power.
The vast majority of people can't do it on the first or second try, they'll just meekly press on the brake like they're coasting to a stop. After more coaching that hard means hard, they start to get it, but it takes many many tries.
2 replies →
The reason attentive humans don't equal the Waymo here is reaction time. When a thing happens the human takes a moment to process what it means, and choose a reaction. It's not, by our standards, a long time but it's way longer than it takes the Waymo.
Going early means you slow early, which means you also take longer to reach the child, but you're braking for all of that extra time, so you're slowing down even more.
It would be nice to see the video (although maybe there are some privacy issues, it is at a school after all).
Anyway, from the article,
> According to the NHTSA, the accident occurred “within two blocks” of the elementary school “during normal school drop off hours.” The safety regulator said “there were other children, a crossing guard, and several double-parked vehicles in the vicinity.”
So I mean, it is hard to speculate. Probably Waymo was being reasonably prudent. But we should note that this description isn’t incompatible with being literally in an area where the kids are leaving their parents’ cars (the presence of “several double parked cars brings this to mind). If that’s the case, it might make sense to consider an even-safer mode for active student unloading areas. This seems like the sort of social context that humans might have and cars might be missing.
But things speculation. It would be good to see a video.
> But most humans would have been aware of the big picture scenario much earlier. Are there muliple kids milling around on the sidewalk? Near a school? Is there a big truck/SUV parked there?
Waymos do this and have for years. They know where the people are around them and will take precautionary action based on that.
Here's a video from 2019 of one understanding that a car in the bike lane means the cyclists may dart out into the lane it's in and taking action based on that. https://waymo.com/blog/2019/05/safety-at-waymo-self-driving-...
That video is nearly 7 years old at this point and they've gotten much, much better since then.
If you think a fully-attentive human driver would have done better, I think you're kidding yourself.
I know you didn't make this point, but if anyone think the average LA driver would have done better than this I've got a bridge to sell you and that's really what matters more. (I say that as someone who used to live like half a mile from where this happened)
> The car only slows down after seeing someone.
How do you know that? The article says it slowed from 17 mph. That’s cautious progress speed, not cruising speed.
In principle, attentive drivers, who have either somehow come independently to the appropriate understanding or have been trained in how to react to hazards ahead...
https://www.gov.uk/theory-test/hazard-perception-test
... could in some circumstances know that there's a likelihood that a child will emerge suddenly and reduce their speed in anticipation where circumstances allow.
Note that: If you cut speed but other drivers can't see why they may overtake, even unsafely, because you are a nuisance to them. Slowing in anticipation that a child will run out from behind the SUV, only for a car behind you to accelerate around you and smack straight into the child at even higher speed, is not the desired outcome even though you didn't hurt anybody...
And yes, we'd need to see the video to know. It's like that Sully scenario. In a prepared test skilled pilots were indeed able to divert and land, but Sully wasn't prepared for a test. You're trained to expect engine failure in an aeroplane - it will happen sometimes so you must assume that, but for a jet liner you don't anticipate losing both engines, that doesn't happen. There's "Obviously that child is going in the road" and "Where the fuck did they come from?" and a lot in between and we're unlikely to ever know for sure.
I live in an area where there are pedestrians stepping into the street without looking, all over the place, and you can drive / cycle without hitting them but have to slow down appropriately if you have to go near something that you can't see behind. Like you say it would be interesting to see the video.
It was already moving slowly. 17MPH is pretty conservative. Most human drivers going past my local school are doing at least 30.
There's a bus stop right behind my house. I routinely hear the driver honking and yelling at people who ignore when the stop sign is extended (which is a misdemeanor in my state). So forgive me for not assuming a human would have done better.
The car was driving 17mph before braking. I don’t think I’ve ever seen a human drive at 17mph in a school zone or other area children congregate.
Meaning you’ve never seen a human drive that slowly in such an area, or you've never seen a human exceed the speed limit in a school zone?
1 reply →
I drive like this too, but I think we’re a small minority. Especially here in LA.
Are they not using a ton of ML to take exactly this sort of context into account?
> But most humans would have been aware of the big picture scenario much earlier. Are there muliple kids milling around on the sidewalk? Near a school? Is there a big truck/SUV parked there?
Waymos constantly track pedestrians nearby, you can see it on the status screen if you ride in one. So it would be both better able to find pedestrians and react as soon as one was on a collision course. They have a bit more visibility than humans do due to the sensor placement, so they also can see things that aren't that visible to a person inside the car, not to mention being constantly aware of all 360 degrees.
While I suppose that in theory, a sufficiently paranoid human might outdo the robot, it looks to me like it's already well above the median here.
Do they speculate about things like “we’re near a school zone, kids are unloading, there might be a kid I’ve never seen behind that SUV?” (I’m legitimately asking I’ve never been in a Waymo).
4 replies →
>The car can react faster that I can after seeing someone
and that can potentially allow internal planning algorithm to choose more risky and aggressive trajectories/behavior, etc. say to reach target destination faster and thus deliver higher satisfaction to the passengers.
Anecdote, but I live next to an elementary school and also on a route frequented by Waymos. Human drivers routinely cruise down the 25mph roads at 40+ and blow stop signs, even during school intake and release. Waymo vehicles always seem a lot more cautious.
When thinking about these things you have to factor in the prior probability that a driver is fully attentive, not just assume they are.
If you’ve ever been in a Waymo you quickly realize their field of view is pretty good. You often see the vehicle sensing small pets and children that are occluded to a passenger or driver. For this reason and my experience with humans near aforementioned school, I doubt a human would out perform the Waymo in this particular incident and it’s debatable they even have more context to inform their decisions.
All that said, despite having many hours in a Waymo, it’s not at all clear to me how they factor in sidewalk context. You get the sense that pedestrians movement vectors are accounted for near intersections, but I can’t say I’ve experienced something like a slow down when throngs of people are about.
Precisely. Environmental context is not considered in Waymo's "peer-reviewed model" (I encourage reflexive commenters to first read it: https://waymo.com/safety/collision-avoidance-benchmarking), only basic driver behavior and traffic signal timings.
Note the weaselly "immediately detected the individual as soon as they began to emerge" in the puff piece from Waymo Comms. No indication that they intend to account for environmental context going forward.
If they already do this, why isn't it factored in the model?
How is "immediately detected the individual as soon as they began to emerge" worded weaselly?
1 reply →
Possibly, but Waymos have recently been much more aggressive about blowing through situations where human drivers can (and generally do) slow down. As a motorcyclist, I've had some close calls with Waymos driving on the wrong side of the road recently, and I had a Waymo cut in front of my car at a one-way stop (t intersection) recently when it had been tangled up with a Rivian trying to turn into the narrow street it was coming out of. I had to ABS brake to avoid an accident.
Most human drivers (not all) know to nose out carefully rather than to gun it in that situation.
So, while I'm very supportive of where Waymo is trying to go for transport, we should be constructively critical and not just assume that humans would have been in the same situation if driving defensively.
Certainly, I'm not against constructive criticism of Waymo. I just think it's important to consider the counterfactual. You're right too that an especially prudent human driver may have avoided the scenario altogether, and Waymo should strive to be that defensive.
> I'm not against constructive criticism of Waymo.
I feel like you have to say this out loud because many people in these discussions don't share this view. Billion dollar corporate experiments conducted in public are sacrosanct for some reason.
> I just think it's important to consider the counterfactual
More than 50% of roadway fatalities involve drugs or alcohol. If you want to spend your efforts improving safety _anywhere_ it's right here. Self driving cars do not stand a chance of improving outcomes as much as sensible policy does. Europe leads the US here by a wide margin.
16 replies →
Absolutely, I can tell you right now that many human drivers are probably safer than the Waymo, because they would have slowed down even more and/or stayed further from the parked cars outside a school; they might have even seen the kid earlier in e.g. a reflection than the Waymo could see.
It seems it was driving pretty slow (17MPH) and they do tend to put in a pretty big gap to the right side when they can.
There are kinds of human sensing that are better when humans are maximally attentive (seeing through windows/reflections). But there's also the seeing-in-all-directions, radar, superhuman reaction time, etc, on the side of the Waymo.
2 replies →
[dead]
I think my problem is that it reacted after seeing the child step out from behind the SUV.
An excellent driver would have already seen that possible scenario and would have already slowed to 10 MPH or less to begin with.
(It's how I taught my daughter's to drive "defensively"—look for "red flags" and be prepared for the worst-case scenario. SUV near a school and I cannot see behind it? Red flag—slow the fuck down.)
First, it's still the automobile's fault.
At least it was already slowed down to 17 mph to start. Remember that viral video of some Australian in a pickup ragdolling a girl across the road? Most every comment is "well he was going the speed limit no fault for him!" No asshole, you hit someone. It's your fault. He got zero charges and the girl was seriously injured.
You seem to be implying that there are no circumstances in which a vehicle can hit a pedestrian and the driver not be at fault... which is absurd.
8 replies →
You mean the Aussie one where the guy was going an appropriate speed for the area and when the cops arrived the parents and their neighbors LIED TO THE POLICE and said he was hooning down the road at excess speed and hit the kid? And that he was only saved from prison by having a dash cam that proved the lies to be lies? That one?
That logic is utter bs, if someone jumps out when you're travelling at an appropriate speed and you do your best to stop then that's all that can be done. Otherwise by your logic the only safe speed is 0.
1 reply →
That’s not how fault works
It's not the drivers fault when they hit a kid who darts out in front of them and they had no time to react and weren't doing anything illegal like speeding.
1 reply →
Aye, and to always look for feet under and by the front wheel of vehicles like that.
Stopped buses similarly, people get off the bus, whip around the front of them and straight into the streets, so many times I’ve spotted someone’s feet under the front before they come around and into the street.
Not to take away from Waymo here, agree with thread sentiment that they seem to have acted exemplary
You can spot someone's feet under the width of a bus when they're on the opposite side of the bus and you're sitting in a vehicle at a much higher position on the opposite side that the bus is on? That's physically impossible.
3 replies →
I don't see how that's feasible without introducing a lot of friction.
Near my house, almost the entire trip from the freeway to my house is via a single lane with parked cars on the side. I would have to drive 10 MPH the entire way (speed limit is 25, so 2.5x as long).
Why can't we add friction to save lives? Automobiles are the single leading cause of death for children in the USA! We're not talking about something uncommon.
Remove the free parking if that's making the road unsafe. Or drive 10 mph. Done.
But you most likely don't have that entire road be full of little kids in the sidewalk all the way. If you did, then yes probably 10mph or less would be wise.
Yes.
- Parked cars on the street. - Drive somewhat fast. - Avoid killing people.
Pick two.
It's hard to consider it "lots of friction" in a vehicle where you press a button to go faster and another button to slow down.
A single lane residential street with zero visibility seems like an obvious time to slow down. And that's what the Waymo did.
3 replies →
I mean, you are putting your finger right on the answer: the whole car thing doesn't work or make sense, and trying to make autonomous vehicles solve the unsolvable is never going to succeed.
1 reply →
>reacted after seeing the child step out from behind the SUV.
Lmao most drivers I see on the roads aren't even capable of slowing down for a pedestrian crossing when the view of the second half of the crossing is blocked by traffic (ie they cannot see if someone is about to step out, especially a child).
Humans are utterly terrible drivers.
They don't even stop when it's a crosswalk with a flashing light system installed and there are no obstructions.
>> Humans are utterly terrible drivers
Duh, driver is, essentially, a type of specialized profession. It's kinda unreasonable to think that everyone could learn to do it well
Good thing we have public transport! :D
Yes and no. Tons of situations where this is simply not possible, whole traffic goes full allowed speed next to row of parked cars. If somebody unexpectedly pops up distracted, its a tragedy guaranteed regardless of driver's skills and experience.
In low traffic of course it can be different. But its unrealistic to expect anybody to drive in expectation that behind every single car passed there may be a child jumping right in front of the car. That can be easily thousands of cars, every day, whole life. Impossible.
We don't read about 99.9% of the cases where even semi decent driver can handle it safely, but rare cases make the news.
I slow down considerably near parked cars. And I try to slow down much earlier approaching intersections where there are parked cars blocking my view of cross walk entries. I need to be able to come to full stop earlier than intersection if there happens to be a pedestrian there.
I kind of drive that way. I slow down, move as far away in my lane from the parked cars as possible. It's certainly what I would expect from a machine that would claim to be as good as the best human driver.
7 replies →
This is generally the problem with self-driving cars, at least in my experience (Tesla FSD).
They don't look far enough ahead to anticipate what might happen and already put themselves in a position to prepare for that possibility. I'm not sure they benefit from accumulated knowledge? (Maybe Waymo does, that's an interesting question.) I.e., I know that my son's elementary school is around the corner so as I turn I'm already anticipating the school zone (that starts a block away) rather than only detecting it once I've made the turn.
Tesla FSD is leagues behind Waymo; generalizing based on your Tesla experience doesn't make sense.
3 replies →
Yes I agree, but why 10mph? Why not 5mph? or 2mph? You'll still hit them if they step out right in front of you and you don't have time to react.
Obviously the distances are different at that speed, but if the person steps out so close that you cannot react in time, you're fucked at any speed.
10mph will do serious damage still, so please for the sake of the children please slow yourself and your daughter's driving down to 0.5mph where there are pedestrians or parked cars.
But seriously I think you'd be more safe to both slow down and also to put more space between the parked cars and your car so that you are not scooting along with a 30cm of clearance - move out and leave lots of space so there is more space for sight-lines for both you and pedestrians.
multiple children in my area have died due to being hit by distracted drivers driving near schools. One incident resulted in 2 children being dragged 60 yards. Here's a snippet from an article about the death I was referencing:
> The woman told police she was “eating yogurt” before she turned onto the road and that she was late for an appointment. She said she handed her phone to her son and asked him to make a call “but could not remember if she had held it so face recognition could … open the phone,” according to the probable cause statement.
> The police investigation found that she was traveling 50 mph in a 40 mph zone when she hit the boys. She told police she didn’t realize she had hit anything until she saw the boys in her rearview mirror.
The Waymo report is being generous in comparing to a fully-attentive driver. I'm a bit annoyed at the headline choice here (from OP and the original journalist) as it is fully burying the lede.
I usually take extra care when going through a school zone, especially when I see some obstruction ('behind a tall SUV', was the waymo overtaking?), and overtaking is something I would probably never do (and should be banned in school zones by road signs).
This is a context that humans automatically have and consider. I'm sure Waymo engineers can mark spots on the map where the car needs to drive very conservatively.
> especially when I see some obstruction ('behind a tall SUV', was the waymo overtaking?)
Yep. Driving safe isn't just about paying attention to what you can see, but also paying attention to what you can't see. Being always vigilant and aware of things like "I can't see behind that truck."
Honestly I don't think sensor-first approaches are cut out to tackle this; it probably requires something more akin to AGI, to allow inferring possible risks from incomplete or absent data.
I appreciate your sensible driving, but here in the UK, roads outside schools are complete mayhem at dropping off/picking up times. Speeding, overtaking, wild manoeuvres to turn round etc.
When reading the article, my first thought was that only going at 17mph was due to it being a robotaxi whereas UK drivers tend to be strongly opposed to 20mph speed limits outside schools.
Most US states cap speed limits around schools at 15mph when children are present. There may also be blinking lights above these signs during times that will be likely.
I'm not sure how much of that Waymo's cars take into account, as the law technically takes into account line of sight things that a person could see but Waymo's sensors might not, such as children present on a sidewalk.
3 replies →
School pick up and drop off traffic is just about the worst drivers anywhere. Like visibly worse than a bunch of "probably a little drunk" people leaving a sports stadium. It's like everyone reverts to "sixteen year old on first day behind the wheel" behavior. It's baffling. And there's always one token dad picking up his kid on a motorcycle or in a box truck or something that they all clutch their pearls at.
It depends. A driver may have seen a child dart behind a car and expect them to emerge on the other side.
Does Waymo have the same object permanence and trajectory prediction (combined) to that of a human?
Once the video evidence it out, it might become evident.
Generally Waymo seems to be a responsible actor so maybe that is the case and this can help demonstrate potential benefits of autonomous vehicles.
Alternatively, if even they can't get this right then it may cast doubts about the maturity of the entire ecosystem
> Does Waymo have the same object permanence and trajectory prediction (combined) to that of a human?
On this note specifically ive actually been impressed, ie when driving down Oak st in SF (fast road, tightly parked cars) I've often observed it slow if someone on a scooter on the sidewalk turns to look toward oncoming traffic (as if to start riding), or to slow passing parked box trucks (which block vision of potential pedestrians)
“Does Waymo have the same object permanence and trajectory prediction (combined) to that of a human?”
Good technical question
If I was a human driver in that contextual situation I wouldn't even be going 14mph in the first place...
>It's likely that a fully-attentive human driver would have done worse.
Maybe. Depends on the position of the sun and shadows, I'm teaching my kids how to drive now and showing them that shadows can reveal human activity that is otherwise hidden by vehicles. I wonder if Waymo or other self-driving picks up on that.
I don't fail to believe that, a child running from behind an suv is really scary
This exact scenario happened with my dad 50 years ago when a little girl ran out to the street from between some parked cars. It's an extremely difficult scenario to avoid an accident in.
A human driver in a school zone during morning drop off would be scanning the sidewalks and paying attention to children that disappear behind a double parked suv or car in the first place, no?
As described by the nhtsa brief:
"within two blocks of a Santa Monica, CA elementary school during normal school drop off hours; that there were other children, a crossing guard, and several double-parked vehicles in the vicinity"
The "that there were other children, a crossing guard, and several double-parked vehicles in the vicinity" means that waymo is driving recklessly by obeying the speed limit here (assuming it was 20mph) in a way that many humans would not.
I live near a school zone in LA and most drivers do not obey school zone speed limits.
You will get honked at by aggro drivers if you slow down to the school zone speed limit of 25mph. Most cars go 40ish.
And ofc a decent chunk of those drivers are on tiktok, tinder, Instagram, etc
Some human drivers? Yes, certainly.
Your median human driver? Sadly, I think not. Most would be rushing, or distracted, or careless.
> waymo is driving recklessly by obeying the speed limit here (assuming it was 20mph) in a way that many humans would not.
I don't think we can say at all that the Waymo was driving recklessly with the data we currently have
> It's likely that a fully-attentive human driver would have done worse.
Why is it likely? Are we taking the vendor's claims in a blog post as truth?
"fully attentive human driver ..." is Waymo's claim, and it could be biased in their favor.
Could be! In aggregate though, Waymos have shown to be safer than human drivers, so my prior is that that holds here.
who benefits from a statement like this?
People who reflexively assume a human driver would do better
It's possible, but likely is a heavy assertion. It's also possible a human driver would have been more aware of children being present on the sidewalk and would have approached more cautiously given obstructed views.
Please please remember that any data from Waymo will inherently support their position and can not be taken at face value. They have significant investment in making this look more favorable for them. They have billions of dollars riding on the appearance of being safe.
I remember someone using similar language when Uber self driving killed someone - and when the video was released, it was laughable.
It is also crazy that this happened 6 days ago at this point and video was NOT part of the press releases. LOL
LOL
OLO
[dead]
I wonder if that is a "fully attentive human drive who drove exactly the same as the Waymo up until the point the child appeared"?
Personally, I slow down and get extra cautious when I know I am near a place where lots of kids are and sight lines are poor. Even if the area is signed for 20 I might only be doing 14 to begin with, and also driving more towards the center of the road if possible with traffic.
I do the same, and try to actively anticipate and avoid situations like this. Sadly, in my experience most drivers instead fixate on getting to their destination as fast as possible.
A fully attentive human would've known he was near a school and wouldn't have been driving at 17 mph to begin with.
Doubt
I said "fully attentive human", not "walking plant".
You clearly don't spend much time around a school measuring the speed of cars. Head on down and see for yourself how often or not a human driver goes >17mph in such a situation.
I said "a fully attentive human". At the very least try to understand my comment before spouting the usual HN snarky reply.
> It's likely that a fully-attentive human driver would have done worse.
> a huge portion of human drivers
What are you basing any of these blind assertions off of? They are not at all born out by the massive amounts of data we have surrounding driving in the US. Of course Waymo is going to sell you a self-serving line but here on Hacker News you should absolutely challenge that. In particular because it's very far out of line with real world data provided by the government.
If you have contradicting data I'd be glad to see it
>It's likely that a fully-attentive human driver would have done worse.
Is based off the source I gave in my comment, the peer-reviewed model
> a huge portion of human drivers
Is based on my experience and bits of data like 30% of fatal accidents involving alcohol
Like I said, if you have better data I'm glad to see it
> based on my experience
The data completely disagrees with you.
> Like I said, if you have better data I'm glad to see it
We all have better data. It's been here the entire time:
https://www.nhtsa.gov/research-data/fatality-analysis-report...
2 replies →
Waymo is intentionally leaving out the following details:
- Their "peer-reviewed model" compares Waymo vehicles against only "Level 0" vehicles. However even my decade-old vehicle is considered "Level 1" because it has an automated emergency braking system. No doubt my Subaru's camera-based EBS performs worse than Waymo's, still it's not being included in their "peer-reviewed model." That comparison is intentionally comparing Waymo performance against the oldest vehicles on the road -- not the majority of cars sold currently.
- This incident happened during school dropoff. There was a double-parked SUV that occluded the view of the student. This crash was the fault of that double-parked driver. But why was the uncrewed Waymo driving at 17 mph to begin with? Do they not have enough situational awareness to slow the f*ck down around dropoff time immediately near an elementary school?
Automotive sensor/control packages are very useful and will be even more useful over time -- but Waymo is intentionally making their current offering look comparatively better than it actually is.
Emergency braking in non-camera/non-LIDAR cars requires a significant radar signal which you're only going to get from another vehicle (and even then it's noisy and tends to produce frustrating false positives, leading to later-than-you-want stops). It very likely won't detect a child or a dog, I'm not aware of a single instance of an EBS claiming to have done so in practice (and kids and dogs get hit every day!).