I was driving across the east side of SF and hit a patch of lights that were out.
The Waymo's were just going really slow through the intersection. It seemed that the "light is out means 4-way stop" dynamic caused them to go into ultra-timid mode. And of course the human drivers did the typical slow and roll, with decent interleaving.
The result was that each Waymo took about 4x as long to get through the intersections. I saw one Waymo get bluffed out of its driving slot by cross traffic for perhaps 8 slots.
This was coupled with the fact that the Waymos seemed to all be following the same route. I saw a line of about a dozen trying to turn left, which is the trickiest thing to navigate.
And of course I saw one driver get pissed off and drive around a Waymo that was advancing slowly, with the predictable result that the Waymo stopped and lost three more slots through the intersection.
On normal days, Waymos are much better at the 4-way stops than they used to be a few years back, by which I mean they are no longer dangerously timid. The Zoox (Amazon) cars are more like the Waymos used to be.
I expect there will be some software tweaks that will improve this situation, both routing around self-induced congestion and reading and crossing streets with dead lights.
Note that I didn't see any actually dead Waymos as others have reported here. I believe this is an extreme failsafe mode, and perhaps related to just too much weirdness for the software to handle.
It would be interesting to see the internal post mortem.
I’m curious what’s the regulation in this scenario? In Canada I think light off means 4-way stop signs so everyone obeys that, or at least most of everyone. What’s the situation in SF?
Yes, that is the same law in California, but so many people drift through stop signs that the guidance is close to meaningless.
In addition, there are 4-way stop signs all over SF and tourists regularly comment on how they work here.
The law is clear - yield to the right, but that is a pretty slow system in congested roads.
The local custom in SF is that someone is usually obviously first, rightmost, or just most aggressive, and opposing pairs of cars go simultaneously, while being wary about left turns.
Of course pedestrians have right of way in California, so someone in a crosswalk gives implied right of way to the road parallel to the person's crosswalk.
The result is 2x or better throughput, and lots of confused tourists.
So ... with the lights out on a Saturday before Xmas, there was a mess of SF local driving protocol, irritated shoppers, people coming to SF for Xmas parties, and just normal Saturday car and foot traffic.
I thought Waymo did pretty well, but as I said, I didn't see any ones that were dead in the middle of the street..
Is this not how four way stops work everywhere? I live in Kansas and have previously lived in Chicago, and I feel like both places follow this custom. Only thing that’s different is the laws are followed slightly more rigorously in low traffic areas, but the customary rules are definitely still in play.
I got stuck behind a Zoox in SF trying to cross the street from an alley. There was an endless stream of stop & go traffic and the Zoox refused to push itself into traffic, despite other cars deliberately giving it space. I wasn't sure if honking at it would help or hurt the situation.
“Thought of the day, and I wish there were a way to get this to legislators:
Come the next Big One earthquake, all of San Francisco’s emergency services will be blocked by Waymos.”
I’m AMAZED they’re not designed to handle this better. This does indeed seem like a massive problem. “Oops we give up” right when things get the worst? How is this OK?
I’ve been very impressed by Waymo’s more cautious approach. Perhaps they haven’t fully thought through the ramifications of it though.
> Perhaps they haven’t fully thought through the ramifications of it though.
There is a chorus of voices here on HN that have tried to do this openly, obviously, myself included. It seems to be almost universally derided by people who apparently assume that we're just trying to hurt a start up out of anti-environmental sentiment and jealousy.
There are more ways to get "self-driving cars" wrong than there are to get it right. Driving is far more complex than the hackers here on Hacker News seem to want to concede, and even if that wasn't the case, I'm not sure where the sentiment that a trillion dollar corporation is naturally going to implement this system with the best interests of society in hand comes from.
What was the better solution here then? Assuming there's hundreds or thousands of self-driving cars suddenly driving in environment without any traffic lights. In the pictures you can see six Waymo cars at a single intersection. Assuming some of them had passengers should they all try to turn at the intersection anyway, when their LIDAR says the lane is likely free and pull over to the side? Is that the safest option? Should there be human police to direct the self driving cars through intersections? Or wait out the temporary electricity failure?
I believe the answer is far more complicated than it seems and in practice having the cars stay still might have been the safest option any of the parties could agree on (Waymo's office, the city traffic people, state regulators, etc).
There are people thinking this stuff out and those cars can 100% pull over automatically but an explicit choice was made not to do so for safety.
I think part of the problem is they’ve made it our problem.
Look I like Waymo. I think they’re neat and I trust them far more than any of the other companies. But in my mind being able to handle stuff like this is just a requirement to be on the roads in any non-trivial number. Like if they had two vehicles in this happened then OK that’s a problem but it was two vehicles in an entire city.
When you have enough on the road that you can randomly have six at one intersection you should absolutely be able to handle this by then.
I want them to do good. I want them to succeed. But just like airliners this is the kind of thing where people’s safety comes first.
What we saw happen looks like the safety of the Waymo and its passengers came above everyone else despite having no need to do that. There are certainly some situations where just staying put is the best decision.
The power went out and there are no other hazards on the road is not one of them. They made things worse for everyone else on average in a foreseeable situation where it was totally unnecessary. And that’s not OK with me.
This feels like the kind of thing that absolutely should’ve been tested extremely well by now. Before they were allowed to drive in large volumes.
I live in the affected neighborhood. There were hundreds of drivers that did not know how to handle a power outage... it was a minority of drivers, but it was a nontrivial, but nominally large number. I even saw a Muni bus blow through a blacked out intersection. The difference is the Waymos failed in a way that prevented potential injury, whereas the humans who failed, all fail in a way that would create potential injury.
I wish the Waymos handled it better, yes, but I think that the failure state they took is preferable to the alternative.
Locking down the roads creates a lot of potential injuries too.
And "don't blow through an intersection with dead lights" is super easy to program. That's not enough for me to forgive them of all that much misbehavior.
Specifically identifying road signs, traffic lights, and dead traffic lights is a narrow problem that has feasible solutions. To the point where we can reasonably say “yeah, this sub-component basically works perfectly.”
Compared to the overall self-driving problem which is very much not a super easy problem.
Right. You know there are humans somewhere in the city who got confused or scared and mess up too. Maybe a young driver who is barely confident in the first place on a temporary permit, or just someone who doesn’t remember what you do and was already over-stressed.
Whatever, it happens.
This was a (totally unintentional) coordinated screw up causing problems all over as opposed to one small spot.
Yeah, the correlated risk with AVs is a pretty serious concern. And not just in emergencies where they can easily DDOS the roads, but even things like widespread weaknesses or edge cases in their perception models can cause really weird and disturbing outcomes.
Imagine a model that works real well for detecting cars and adults but routinely misses children; you could end up with cars that are 1/10th as deadly to adults but 2x as deadly to children. Yes, in this hypothetical it saves lives overall, but is it actually a societal good? In some ways yes, in some ways it should never be allowed on any roads at all. It’s one of the reasons aggregated metrics on safety are so important to scrutinize.
We already have a solution, it's written down in the traffic laws. If the signals fail, treat the intersection roughly like a four-way stop. Everybody learns this in drivers' ed. It's not obscure. If the cars can't follow traffic rules maybe they're not ready to be on the streets unsupervised.
That may be the rules for humans, particuarly people who are always in a rush and won't stay still anyway. With a major intersection turned four-way stop you have lots of humans making very complex decisions and taking a lot of personal risk. If multiple self driving cars make the choice at the wrong time you could jam up an intersection and create a worse traffic issue, or kill a passenger.
It's all a careful risk calculation, those self driving cars need to determine if it's safe to continue through an intersection without the traffic lights their computers spent millions of hours to train on (likewise with humans). That's a tough choice for a highly regulated/insured company running thousands of cars.
If anything, their programming should only take such a risk to move out of the way for a fire truck/ambulance.
The problem seems to be that the Waymo cars did exactly as you requested and treated the intersections like 4 way stops but kept getting displaced by more aggressive drivers who simply slowed and rolled.
How many non-Waymo accidents happened at intersections during this time? I suspect more than zero given my experiences with other drivers when traffic lights go off. Apparently, Waymo's numbers are zero so humans are gonna lose this one.
The problem here is that safety and throughput are at odds. Waymo chose safety while most drivers chose throughput. Had Waymo been more aggressive and gotten into an accident because it wouldn't give way, we'd have headlines about that, too.
The biggest obstacle to self-driving is the fact that a lot of driving consists of knowing when to break the law.
> Did they? They chose their safety. I suspect the net effect of their behavior made the safety of everyone worse.
There is no viable choice other than prioritizing the safety of your rider. Anything less would be grounds for both lawsuits and reputational death.
The fact that everybody else chose throughput over safety is not the fault of Waymo.
Will you also complain when enough Waymo cars start running on the freeways that a couple of them in a row can effectively enforce following distances and speed limits, for example?
Something I had pounded into me when I drove too slowly and cautiously during my first driving test, and failed.
Those Waymos weren't moving which is a pretty egregious example of obstructing traffic.
An old rule of thumb is every time a service expands by an order of magnitude there are new problems to solve. I suspect and hope this is just Waymo getting to one of those points with new problems to solve, and they will find a way to more graciously handle this in the future.
> Will you also complain when enough Waymo cars start running on the freeways that a couple of them in a row can effectively enforce following distances and speed limits, for example?
In my state, that would itself be a traffic violation, so yes I would. The leftmost lane on an interstate highway is reserved for passing. An autonomous vehicle cruising in that lane (regardless of speed) would therefore be programmed in a way that deliberately violates this law.
Enforcement is its own challenge, whether robots or humans.
> The leftmost lane on an interstate highway is reserved for passing.
Sadly, in most states, this is not true anymore. Most of those laws have been repealed.
I was very pleasantly surprised when I was in Colorado that they had explicit signs saying that if you had 5 (I think) or more cars behind you that you were supposed to pull right and let them pass.
However, I wasn't really thinking about a Waymo cruising in the left lane but simply 4 or 5 Waymo's in the right lane going right at the speed limit with proper following distance. That's going to effectively lock the right lane to the speed limit which then means that even a single other car would lock the left lane to the speed limit as well. Basically, even a couple of Waymos in the right lane would drop freeway speeds dramatically.
> Assuming there's hundreds or thousands of self-driving cars suddenly driving in environment without any traffic lights.
Self-driving cars should (1) know how to handle stops, and (2) know that the rules for a failed traffic light (or one flashing red) are those for an all-way stop.
Just pulling over and getting out of the way really would help. There's no reason a human couldn't do the same safely. Not beta testing your cards on public roads would really be ideal. Especially without human drivers ready to take over.
That’s what they usually do. The assumption here is that due to the blackout or some other related issue the human drivers were unavailable.
However even if that’s not true if they have more cars than human drivers there’s gonna be a problem until they work through the queue. And the bigger that ratio, the longer it will take.
I guess that in a blackout they should just have the cars park somewhere safely. Maybe it'd be best to never have more cars on the road than assisting/available human drivers. As soon as no human drivers are available to take over for outage/staffing/whatever reason all cars should just pull over and stop.
> There is a chorus of voices here on HN that have tried to do this openly, obviously, myself included.
Maybe I'm reading things wrong, but it sounds like the top comment wants waymo to be better, and you want waymo to be off the roads. You're not talking about the same kind of "thinking through the ramifications".
> I'm not sure where the sentiment that a trillion dollar corporation is naturally going to implement this system with the best interests of society in hand comes from
The sentiment comes from the corporation itself. With this much money at stake you know they have a hand in steering the conversation and that includes on sites such as this.
Waymo may discover that heavy equipment (large fire trucks can easily push Waymo out of the way if it can find somewhere to push it to) WILL move the cars (at least if there is no one in them at the time) in such cases. I recall the scenes during recent wildfires where abandoned cars were blocking roads and a skip loader was just picking up the cars and dumping/pushing them to the side of the road/over the edge - causing extensive damage to some of them.
Decades ago I recall talking to a fireman expressing a question of what happened if there was a car blocking their access in an emergency and he made it clear that the bumper on the front of the truck and the truck's healthy diesel engine would usually take care of the problem very quickly.
I wouldn't cry for waymo if a bunch of their cars got bulldozed out of the way but that's still unacceptable since it slows down emergency vehicles and first responders.
>I’m AMAZED they’re not designed to handle this better.
This has been the MO for "tech companies" for the past 20 years. Meanwhile I'm told I'm paranoid when the industry of "move fast or break things" decides to move into mission/safety critical industries and use its massive wealth to lobby for deregulation to maintain its habits.
We certainly have BS regulations done to constrain competition. But I'd wager a good 80% of them exist for good reason.
In cases where the traffic signal is not working, it is known that the FSD has to take on a more challenging role of reading traffic agent gestures. I think they have that functionality built in. But not when neither traffic signal is working nor traffic agent is present.
The basic thing is to treat everything like a four-way stop sign.
If we reach the point of needing to forcicbly move mass numbers of cars off the road for fire trucks, that's a dire situation where routine cost/benefit analysis has already gone out the window.
No I won't be happy about it, but yes if I block emergency services and they need to damage my property then I am absolutely the one who should have to pay.
Although it would be amusing for them to do it at high velocity if the cars (and surrounding cars if any) were "dead heading" or had no humans in them for other reasons (perhaps because the humans had fled the vehicles upon seeing the fire truck headed their way!).
Yes that’s the correct decision when those are the only options, like if a car has stalled or the driver just got out and ran away.
In this case there’s a third option: the computer that’s still perfectly functional should have been able to get out of the way on its own. And legally all drivers are required to.
I assume that applies to robots as well, if it doesn’t it absolutely should.
Surely if the Big One hits then all of the metropolitan areas on the West Coast would be gridlocked in scenes reminiscent of zombie apocalypse movies anyway? I guess we won't know until it happens for sure, but I can't imagine it would be easy for emergency services to get around with or without Waymo.
It’s not gonna be good. But you want it to be a gridlock because the cars can’t get out fast enough because there’s too many cars on the road.
Not because a bunch of cars that are perfectly capable of moving are just sitting there blocking things purposefully waiting for the driver in the sky to take over.
And what if, due to $BIG_DISASTER they won’t be able to for a week?
In case of a natural disaster, it’s guaranteed that human drivers will abandon their cars on the road and cause gridlock. It happens all the time. Emergency vehicles are built to handle it.
Because it’s a power outage. If we instead learned about this during a real disaster people could have died because these things were let on the road without planning what they should do in abnormal circumstances.
> If we instead learned about this during a real disaster people could have died
This is universally true. The question is how bad could it have been, and in which cases would it have been the worst?
> We’re lucky it’s not a disaster
This is always true. Again, the question is how lucky?
We have an opportunity to count blocked emergency vehicles and calculate a hypothetical body count. This lets us characterize the damage. But it also permits constraining hysteria.
I'm sure that if this was something predictable like a cyclone or wildfire, Waymo would still have 100% of their nightly traffic on the road, right? And SFFD would not be able to do what they normally do when they can't get support, which is hop into the car and use the controls to manually move it?
Or... maybe Waymo HAS considered what their cars should do in abnormal circumstances and this kind of outcome was considered acceptable for the number of cars and the nature of the "disaster"?
In this country, if heart disease or cancer doesn't kill you, a car probably did.
Until "Waymos lessons" are killing people at that rate, I am 100% OK with a Waymo making my trips an extra 5 minutes longer every 50th trip or whatever else the real stat is.
I was curious if Waymo has even been involved with a crash that killed someone, so I looked it up. The answer is yes - there was a Tesla going 98mph in SoMa whose driver died after hitting a Waymo. Clearly the Waymo's fault!
When we learn our lesson that letting companies beta test on public roads consequence free is just another cost to the rest of us so that a small number of people can enrich themselves at our expense.
Whenever they become so much a problem that they counterbalance public and private interests in having and improving robotaxis. For most people, we are nowhere near that.
They are. I did myself yesterday because one was sitting at the front of a turning lane at a dead light, just waiting there forever with the blinker on.
It can do that with a normal driver too. Doesn’t make it ok for there to ever be a situation where they need to when the target vehicle/driver is just fine and capable of doing it themselves.
I'm surprised the don't know to treat it as a 4-way stop, either. This kind of outage is pretty common in Phoenix, too, which is another major Waymo market. It probably happens to at least some part of the city every monsoon season.
What if there was a herd of people off-shore on-call willing to basically "RDP in" and take over control (human takeover) of the entire fleet when needed? I could see that being an attractive pitch.
Latency makes this hard even with local connections, it’s essentially impossible due to physics to do it offshore.
And I believe Waymo remote access only allows providing high level instructions (like pull over, take the next right, go around this car, etc) precisely because full direct control with a highly and variably latent system is very hard/dangerous.
And in an emergency situation you’re likely to have terrible connectivity AND high level commands are unlikely to be sufficient for the complexity of the situation.
I suspect that in a large scale disaster/emergency the communications systems may be disrupted and it may not be possible to remotely control the vehicles.
Perhaps in such cases they can pull over in a safe place, or if they have an occupant ask them if they wish to continue the journey or stop.
Perhaps they already do this, I have no experience with autonomous vehicles.
But even without them getting better, as far as I know there were zero waymo fatalities due to this.
That's more than I can say about Helene, where there was at least one fatality due to traffic light outages.
Lets not forget that a big part of why we want Waymo is that it has already lead to a dramatic decrease in fatal accidents. They are a great company that will do a lot of good for the world. One bad night (in which noone was hurt, in part because of their cautiuosness) shouldn't negate that.
Seeing as how literally nobody died, I'm not sure if I agree with your sentiment.
I was curious if Waymo has even been involved with a crash that killed someone, so I looked it up. The answer is yes - there was a Tesla going 98mph in SoMa whose driver died after hitting a Waymo. Clearly we should shut down Waymo until they can handle that situation!
You're being sarcastic, but it's a valid point. I'd love to know if there were any traffic fatalities at all during the affected period. Chances are there were and that they were due to human error.
In my experience, humans respond incredibly poorly to traffic lights being out. There's no sense or reason, just people deciding to drive across the intersection when they feel like it's okay.
Presumably Waymo will make sure they can handle this situation in the future, but I'm not sure there's a really satisfactory solution. The way you're supposed to handle an intersection with no lights (treat it as a stop sign intersection) doesn't work very well when no one else is behaving that way.
That wasn’t my experience, having just driven across the city and back during tonight’s outage. It was actually weirdly inspiring how well people coordinated at so many of the powerless intersections.
There was a lot of confusion, and some people took advantage of it to rush through, but for the most part it was pretty orderly. Which makes sense because in many parts of the world where there are no traffic lights or stop signs, people get on just fine.
The Waymo’s, on the other hand, were dropping like flies. While walking from Lower to Upper Haight I spotted a broken Waymo every handful of blocks. The corner of Haight & Fillmore was particularly bad, with 3 of them blocking traffic in both directions — in the path of both the 7 and 22 bus lines.
There have been some experiments that suggest that traffic flow is both safer and more efficient if you just turn disable the signals entirely. I doubt it applies to all, or even most, situations but it's definitely food for thought: https://www.npr.org/2008/01/19/18217318/german-towns-traffic...
I think a significant factor helping that to work is the mixing of all traffic on the street. I've noticed that in LA's Skid Row, where homeless people are constantly moving into the street on foot or on bicycle and they walk around in vehicle lanes pushing shopping cart armadas and so on, drivers are more cautious than usual and I see, if anything, less reckless driving and close calls there than in other parts of downtown, where pedestrians stick to the sidewalk and distracted or car-brained drivers don't look out for them. Just anecdotal observation, of course.
Different things. A country with lax rules is not the same as a specific environment with shared spaces, where according to known data it's safer to eliminate some specific kind of regulation and let the remaining part take over.
My own experience has given me a somewhat more-nuanced take.
At first, it's akin to the path of evil. Way too many people just zoom through intersections with dark traffic lights like they're cruising unimpeded down the Interstate, obvious to their surroundings. Some people get grumpy and lay on the horn as if to motivate those in front of them to fly through themselves.
But many people do stop, observe, and proceed when it is both appropriate and safe.
After awhile, it calms down substantially. The local municipality rounds up enough stop signs to plant in the middle of the intersections that people seem to actually be learning what to do (as unlikely as that sounds).
By day 2 or 3, it's still somewhat chaotic -- but it seems "safe" in that the majority of the people understand what to do (it's just stop sign -- it may be a stop sign at an amazingly-complex intersection, but it's still just a stop sign) and the flyers are infrequent-enough to look out for.
By day 5 or 6, traffic flows more-or-less fine and it feels like the traffic lights were never necessary to begin with. People stop. They take turns. They use their turn signals like their lives depend on it. And the flyers apparently have flown off to somewhere else. It seems impossible to behold, but I've seen it.
But SF's outage seems likely to be a lot shorter than that timeline, and I definitely agree with Waymo taking the cautious route.
(but I also see reports that they just left these cars in the middle of the road. That's NFG.)
Huh, here in Germany we have street signs (mostly of a "you are the priority road" 45° rotated square yellow-on-white "sunny side up egg" sign and the "you are not the priority road" down-pointing white-on-red triangle; for 3-way if the priority road isn't the straight road or the concept of straight is ambiguous, there's a supplemental sign depicting the path of the priority road) permanently on traffic lights; it's also common enough for non-major roads to have the lights turned off at night so drivers tend to be familiar with falling back to the signs when the lights are off.
In absence of priority roads there is also the "right before left" rule which means that the car coming from the right if they would conflict in time is the car that has priority.
It's also always illegal to enter an intersection if you can't immediately clear it; that seems to work better when there are no green traffic lights to suggest an explicit allowance to drive, though.
Sure, but we're pretty far from Germany over here.
In the States (or at least, every US state that I'm familiar with -- each one is free to make their own traffic rules, similar to how each EU member state also has their own regulatory freedoms), a dark/disabled/non-working traffic light is to be treated as stop sign.
For all drivers, in all directions of travel: It functionally becomes a stop sign.
That doesn't mean that it is the best way, nor does it mean that it is the worst way. It simply is the way that it is.
How does "you can only piss with the cock you've got" translate to German slang?
Treating it like a stop sign also doesn't work very well when there are huge amounts of pedestrians. As a pedestrian that yesterday meant I got the right of way all the time. For cars it was mayhem downtown.
In contrast many years ago I lived at an intersection that had almost no pedestrians back then and a few times for a power outage limited to our building and that intersection. I enjoyed standing on my balcony and watch traffic. It mostly worked well. Cars did treat it like a intersection with stop signs. There two issues happened though. One was when there was no car already stopped and about 10%-20% of drivers didn't realize there was an intersection with lights out and just raced through it. The other ironically were bicyclists. 90% of the just totally ignored there was an intersection. That was especially scary when they arrived at the same time as one of those cars who didn't realize it either.
> Treating it like a stop sign also doesn't work very well when there are huge amounts of pedestrians. As a pedestrian that yesterday meant I got the right of way all the time. For cars it was mayhem downtown.
I’ve long been curious if people in S.F. who are used to Waymo “behavior” – including myself – behave differently when a Waymo is involved. For the most part, Waymos are extremely predictable and if you’re on the road as a pedestrian, cyclist, or car, are you more aggressive and willing to assert your turn in the road? Curious if Waymo has done a psychological study of how we start to think about the vehicles. I know many runners, for example, that’ll stride right into the crosswalk in front of a Waymo but wouldn’t dare to do that in front of other cars. Similarly, anecdotally people treat public transit vehicles differently than a civilian car: folks are more willing to let them drive even if out of natural turn.
I saw this recently when the lights were out at an intersection in Manhattan. People kept on driving and almost hitting pedestrians and cars. I called 911 and then directed traffic for 15 minutes until DoT came out and put up a temporary stop sign.
Same setup in the Netherlands, there are right of way signs everywhere that apply when the lights don't work.
One interesting effect is that there are also often pedestrian crossings that have priority over everyone. Normally those are limited by lights, but without lights a steady stream of pedestrians stops all traffic. Seen that happen in Utrecht near the train station recently, unlimited pedestrians and bikes, so traffic got completely stuck until the police showed up.
Except in NL you guys actually have someone sit and think about the flow of traffic, in most other countries the design is usually pretty much random. In my parents' city in Poland they decided to turn off the traffic lights at night and the result is much higher number of accidents. Funnily, my father failed to yield the right of way exactly when talking about the issue. Yes, theoretically all intersections do have full set of signs, but in practice the visibility of those signs is extremely limited.
> Same setup in the Netherlands, there are right of way signs everywhere that apply when the lights don't work.
Most places in Poland have this exact setup. And I say most because I have not seen one that does not, but I am guessing they exist. Maybe some of them have bad visibility even?
If one does not respect the yield sign that does not seem a signaling problem.
Some don't in Finland. But then it is back to basic rule of yielding to traffic from right. Which is pretty common on smaller roads so drivers should think about it enough.
The words sum up to "don't go straight unless you have business being there" but still, it really sums up the approach to signaling. Nobody in the whole command chain sat and thought "that might be a tiny bit difficult to parse".
I mostly read those signs as “don't go there unless you already know you can” which as a “tourist” I just assume can't (unless I'm a local, and figure it out).
What I've recently found troubling is the places that use similar signs for emissions controls. With a rental you usually have a recent enough car that you can ignore those.
Being able to distinguish between “low emissions zone, but any car from this decade can go in” and “local traffic only, you need to live in this neighborhood to enter” in a foreign language, bit me a couple of times while traveling.
> In my experience, humans respond incredibly poorly to traffic lights being out.
My purely anectodotal experience is that the response is variable and culturally dependent. Americans tend to treat any intersections with a downed stoplight as a multi-way stop. It's slow but people get through. I've experienced other countries where drivers just proceed into the intersection and honk at each other. (Names withheld to protect the innocent.)
It seems a bit like the Marshmallow test but measures collaboration. [0]
Where I'm from (a relatively rural country) they just get treated like give way signs you'd have on a country road, the larger road generlly has priority but as it's not clear they'd be more cautious too.
One could argue that it's "cultural", but California state law says this about the situation:
> Traffic Light Not Working
When a traffic light is not working, stop as if the intersection is controlled by STOP signs in all directions. Then proceed cautiously when it is safe to do so.
Not really, they just treat it the intersection as if it were ruled by stop signs. It’s not always easy to keep track of who goes next but overall people can handle themselves pretty well in those cases.
Neither a lack of traffic lights nor cell service should cause the Waymos to stop in the middle of the road, that’s really troubling. I can understand the system deciding to pull over at the first safe opportunity, but outright stopping is ridiculous.
AIUI, it was the irregularity of the uncontrolled intersections combining with the “novel” (from the POV of the software) driving style of the humans. In dense areas during outages signaled intersections don’t actually degrade to 4 way stops, drivers act pretty poorly.
The normal order and flow of traffic broke down. The software determined it was now outside its safe parameters and halted.
Certainly not ideal, and the should be a very strong regulatory response (the gov should have shut them down), and meaningful financial penalties (at least for repeat incidents).
Perhaps this is by design. Cruise had a failsafe system that detected a collision and decided to pull over but by pulling over it dragged a person underneath the car (or something close to this scenario). Maybe this dumb failsafe was designed not to repeat Cruise's mistakes?
Certainly a better way to handle this would have been to pull over. I think stopping where ever it happened to be is only acceptable if the majority of sensors fail for some reason
I was there. I encountered multiple stopped Waymos in the street. It was annoying, but not dangerous. They had their lights on. Any driver following the rules of the road would get around them fine. It was definitely imperfect, but safe. Much safer than the humans blowing through those very same intersections.
When I was a young man, I worked at a restaurant, and the lights went off.
I being the hero I was, wanted to keep the show running, bought some candles, ovens worked fine, water worked fine (for now). I wanted to charge cash. But eventually big boss came and shut us down since light wasn't coming.
And he was right, cooking and working under those conditions is dangerous for the staff, but also for the clients, without light you cannot see the food, cannot inspect its state, whether stale, with visible fungi, etc...
Yes, the perfect worker would still operate under those conditions, but we are not perfect, and admitting that we only can provide 2 or 3 nines, and recognizing where we are in that 0.01% moment, is what keeps us from actually failing so catastrophically that we undo all of the progress and benefits that the last bit of availability would have allowed us.
Waymos rely on remote operators to take over when the vehicle doesn't know what to do, and obviously if the remote connection is gone then this is no longer available, and one might speculate that the cars then "fail safe" by not proceeding if they are in a situation where remote help is called for and inaccessible.
Perhaps traffic lights being out is what caused the cars to stop operating autonomously and try to phone home for help, or perhaps losing the connection home is itself enough to trigger a fail safe shutdown mode ?
It reminds a bit of the recent TeslaBot video, another of their teleoperated stunts, where we see the bot appearing to remove a headset with both hands that it wasn't wearing (but that it's remote operator was), then fall over backwards "dead" as the remote operator evidentially clocked off his shift or went for a bathroom break.
That’s clearly unacceptable. It needs to gracefully handle not having that fallback. That is an incredibly obvious possible failure.
Things go wrong -> get human help
Human not available -> just block the road???
How is there not a very basic “pull over and wait” final fallback.
I can get staying put if the car thinks it hit someone or ran over something. But in a situation like this where the problem is fully external it should fall back to “park myself” mode.
> How is there not a very basic “pull over and wait” final fallback
Barring everything else, the proper failsafe for any vehicle should be to stop moving and tell the humans inside to evacuate. This is true for autonomous vehicles as well as manned ones–if you can't figure out how to pull over during a disaster, ditching is absolutely a valid move.
If the alternative is that the vehicle explodes, sure. And since GP did say "final fallback", I suppose you're right. But if the cars are actually reaching that point, they probably shouldn't be on the road in the first place.
The not-quite-final fallback should be to pull over.
Yeah. I wasn’t considering people, just getting the car out of the way.
I wasn’t considering people taking it as a given that any time the car gives up the doors should be unlocked for passengers to leave if they feel it’s safe.
And as a passenger, I’d feel way safer getting out if it pulled over instead of just stopped in the middle of the street and other cars were trying to drive around it.
Yes. OP is inferring Waymo's internal processes from this meltdown. ("Makes me think there are likely other obvious use cases they haven’t thought about proactively either.")
If Waymo literally didn't foresee a blackout, that's a systemic problem. If, on the other hand, there was some weird power and cellular meltdown that coïncided with something else, that's a fixable edge case.
> Yes. OP is inferring Waymo's internal processes from this meltdown. ("Makes me think there are likely other obvious use cases they haven’t thought about proactively either.")
No, I'm not inferring internal processes.
I'm guessing level of critical thinking.
When you are creating autonomous vehicles, one of the things that you want to risk assess and have mitigation for is what you want the vehicles to do in case the systems they depend on fail (e.g. electricity, comms).
Now, it could be that the team has anticipated those things but some other failure in their systems have caused vehicles to stop in the middle of intersections, blocking traffic (as per article).
I'm super curious to learn more about what Waymo encountered and how they plan to up their game.
> I'm not inferring internal processes…I'm guessing level of critical thinking
Genuine question: how do these differ? Isn’t the level of critical thinking of Waymo’s employees internal to it? (What’s the mens rea analogue for a company?)
The "coinciding problems" should be an assumption, not a edge case we reason away. Because black swan events are always going to have cascading issues - a big earthquake means lights out AND cell towers overloaded or out, not to mention debris in streets, etc.
What they need is a "shit is fucked fallback" that cedes control. Maybe there is a special bluetooth command any police or ambulance can send if nearby, like clear the intersection/road.
Or maybe the doors just unlock and any human can physically enter and drive the car up to X distance. To techies and lawyers it may sound impossible, but for normal humans, that certainly sounds better. Like that Mitch Hedberg joke, when an escalator is out of order it becomes stairs. When a Waymo breaks it should become a car.
The Waymos still have all their normal driver controls. There is a process where law enforcement can enter the vehicle, call Waymo and verify their law enforcement status, and then switch the vehicle into manual mode and drive it as normal.
Here is their instructions for law enforcement in the Waymo Emergency Response Guide:
>If Waymo literally didn't foresee a blackout, that's a systemic problem.
I agree with this bit
> If, on the other hand, there was some weird power and cellular meltdown that coïncided with something else, that's a fixable edge case.
This is what I have a problem with. That’s not an edge case. There will always be a weird thing no one programmed for.
Remember a few years ago when a semi truck overturned somewhere and poured slimy eels all over the highway? No one‘s ever gonna program for that.
It doesn’t matter. There has to be an absolute minimum fail safe that can always work if the car is capable of moving safely. The fact that a human driver couldn’t be reached to press a button to say to execute that is not acceptable. Not having the human available is a totally foreseeable problem. It’s Google. They know networks fail.
This isn't to disagree with your overall point about proper emergency mitigation and having humans available.
> Remember a few years ago when a semi truck overturned somewhere and poured slimy eels all over the highway? No one‘s ever gonna program for that.
While the cause is unusual, this is really just three things that everyone absolutely should be programming into their autonomous vehicles: accidents, road debris, and slick conditions.
Certainly. That one was interesting both because of the odd specifics of it and because it made the road more slippery than any normal accident where just a bunch of boxes of random dry goods fell over.
It just happens to make a fantastic example of “thing no one is ever going to foresee“.
If there wasn’t footage how many people would even believe it happened?
A fail-safe is EXACTLY blocking roads at intersections without power, not proceeding through intersections without power. It's much safer to be stopped than to keep going. I honestly wish the humans driving through blacked out intersections without slowing down in my neighborhood last night actually understood this.
It’s not a fail-safe. It’s a different failure mode. Jamming up traffic, including emergency traffic, creates systemic problems.
It’s a bit like designing an electronic lock that can’t be opened if the power goes out. If your recourse to exiting a dangerous situation becomes breaking the door, then the lock is unsafe.
Fail-safe means "in a situation where the function fails, fail in a way that doesn't cause injury" -> the cars didn't know how to proceed, so they stopped, with their lights on, in a way that any attentive driver could safely navigate... which is a failing safe.
The alternative here, is a protocol that obviously hasn't been tested. How on earth are you going to test a Waymo in blackout conditions? I would rather have them just stop, than hope they navigate those untested conditions with vulnerable pedestrians and vehicles acting unpredictable.
Simulate them on a test course? There are absolutely places with street lights and everything that you could test something. Hell since they don’t need to work you can just have some put up in a parking lot to test with. Who cares.
You don’t need to wait for a city blackout to actually test this kind of scenario.
The thing still has cameras. And LIDAR. It should be fully capable of pulling over on its own safely. Why would not having a traffic light prevent that?
Humans are expected to negotiate this. The robots should be too. That’s part of driving. If the lights fail, the law tells you what you’re supposed to do. And it is not stop the intersection.
> Fail-safe means "in a situation where the function fails, fail in a way that doesn't cause injury"
In a very local sense, this is true. In terms of the traffic system, this can create a systemic problem if the stoppage causes a traffic jam that creates problems for emergency vehicles.
Thus it is a _different_ failure mode.
If someone stops in the middle of traffic because they’re lost, their GPS went out, or they realized that they’re unsafe to drive, we don’t celebrate that as the driver entering a fail-safe mode. We call that “bad judgment” and give them a ticket.
If it precipitates a larger problem where lives are lost, they may be in considerable legal or financial trouble.
I don’t see why we should treat Waymo any differently.
Traffic doesn’t cause injury. Why are we concerned about traffic flow in a blackout situation. The cars stopped at intersections, EMS could use the oncoming lanes. I’m not seeing how it’s not a fail safe, you’re describing it as not being fail-ideal, and I would agree.
I'm confused, is your concern that enough Waymos shut themselves down on a one way road, at the same intersection, so as to block the intersection? Yes, I could see that as being a concern. I suspect it would be reported almost immediately, and would be at the top of the list of for the folks at Waymo to address. The cars weren't abandoned. They eventually moved. Though, I suspect they had to be manually driven (virtually or otherwise), out of the way. I can see how this could be a problem, but considering it would likely be at the top of the list of problems for Waymo -- again, during an emergency -- that I suspect it's not a serious concern in the long run.
Would I preferred that they had a light turn on that was flashing "An unknown emergency is occurring, please park me"? Yes, I think that would be a better solution. I would have preferred better performance from Waymo. My entire point here is that I'm happy that in my neighborhood, the Waymos were acting in a fail-safe manner, rather than just winging it.
An intersection without power is supposed to be treated as a 4-way stop. An unfortunately high, nontrivial number of drivers last night were not following that rule.
The outrage people would rightly have at Waymo allowing a number of its vehicles to blow the lights would be huge. People running blacked out lights is unacceptable.
> Once you’re on public roads, you need to ALWAYS fail-safe.
Yes.
> And that means not blocking the road/intersections when something unexpected happens.
No. Fail-operational is not the only allowable fail-safe condition for automobiles. For example, it is acceptable for loss of propulsion to cause stop-in-lane — the alternative would be to require high-availability propulsion systems, or to require drivers to always have enough kinetic energy to coast to side. This just isn’t the case.
One can argue that when operating a fleet with correlated failure modes the rules should change a bit, but that’s a separate topic.
Yeah, I was shocked by this. Blackouts in California aren’t some sort of rare event. I’m primed to expect rolling brownouts/blackouts yearly in the summer.
There were significant power shutdowns in California in 2019 (affecting millions of customers in aggregate); the reason for the shutdowns was different from 2001 (preemptive shutdowns when the risk of downed power lines starting wildfires was thought to be high) but the impact on customers is the same: no power for an extended period.
One usage case that I saw myself is when a vehicle is parked such that it will require the other vehicle to go slighty over the curb, in this case the curb is flat so I assuming the parked driver thought it was okay. Every other human driver did okay, but Waymo just refused to put its wheel on the curb and just got stuck. Video here: https://x.com/aaditya_prakash/status/1989444130238259575?s=2...
It also means that their claims of "autonomy" are fraudulent, like most "self driving" cars. A car which depends on powered infrastructure outside the car to drive is not autonomous.
I think the 100 or so miles I can generally drive in my car while it still has gas is different from my car just stopping suddenly because of a power outage.
We should put self driving cars on tracks so they are always out of the way and have easily predictable behavior. Maybe we can even link the cars together for efficiency or something like that.
You can further optimize the setup by not installing engines/motors in all of them. So maybe you have one car providing locomotion, with the rest following behind and designed for carrying.
And all the power could just come from a few large centralized facilities that are super efficient. We could just use thin strands of metal to get it to the vehicles over head…
Expense is correlative to scale, likely it's cheaper to deploy pantographs than battery factories.
Why did India build a high speed freight corridor with overhead power when they could have used batteries instead? Because the quantity of battery to power the trains doesn't exist, and overhead wires do.
They are good for infrequently used track and places where overhead wires would be in the way, like that very Tesla employee shuttle on it's own track and container ports.
It's not the best way to go for mainline track and not suitable for long distance high speed trains.
At this point overhead cables are likely many orders more expensive.
Buses in Paris run with IIRC 60kWh battery and pantograph charger at every other station. Packs (not cells) recently dropped to below $100 kWh. At $6k thats probably what city pays for couple of replacement seats (gold plating et all).
Sorry if you're playing in to the joke, I can't tell. Streetcars / trams were widely deployed before they were ripped out for the car, driven by lobbyists interested in selling cars. Wondering why no one has bothered with that is starting from a false premise, because people have bothered with that.
I'm well aware that there are multiple options for public transport, but none of them are actually as flexible or as far reaching as cars/taxis. To me this 'but why not trains' on every article about self-driving cars is a tired meme that fails to address that these are not equivalent options. I might as well say "Well, why don't we get rid of these expensive rails and fixed timetables and just lay down some cheap concrete and let people navigate how they want" in just as condescending a tone and be equally as unconvincing.
There are car lobbyists of course but the streetcars in LA at least were put there by housing developments to sell suburban homes before most people owned cars. Once the homes were sold the corporation that built the rails had no incentive to maintain them, and eventually they were spun off and went bankrupt (of course competing with cars didn't help)
Nor did the fixed price controls they were often saddled with. It seems that politicians are congenitally completely incapable of considering inflation indexing as a concept when they are writing laws.
Street cars are a red herring anyway. Because street cars don't maintain anywhere near the same number of routes as free-form roads. It is a routing problem still, and railed vehicles perform much, much worse at it, which is why they need to be time multiplexed with rail schedules.
It's tedious to see these same sarcastic comments on every self driving car story. Yes. Buses and trains exist.
When you link the cars together, they usually switch to a hub that's a 10-15 minute walk from your destination instead of your destination and the compartments are occasionally shared with unstable and violent people, which while possibly "efficient" in some metrics, are downsides that many people would rather avoid. Personal compartments are a real differentiating advantage.
violent unstable people aren't inherent to cities.. they're inherent to places that refuse to spend any money on social work/housing/and enjoy punishing people
Rugged American Individualism and Capitalism doesn't allow us to have things like that. We must always be in our individual bubbles away from the filthy poors.
Prior to reading the article, I assumed Waymos were stuck due to an Internet connectivity issue. However, while the root cause is not explicitly stated, it sounds like the Waymos are “confused” by traffic lights being out.
I wonder how Waymos know that the traffic lights are out.
A human can combine a ton of context clues. Like, "Well, we just had a storm, and it was really windy, and the office buildings are all dark, and that Exxon sign is normally lit up but not right now, and everything seems oddly quiet. Evidently, a power outage is the reason I don't see the traffic light lit up. Also other drivers are going through the intersection one by one, as if they think the light is not working."
It's not enough to just analyze the camera data and see neither green nor yellow nor red. Other things can cause that, like a burned out bulb, a sensor hardware problem, a visual obstruction (bird on a utility cable), or one of those louvers that makes the traffic light visible only from certain specific angles.
Since the rules are different depending on whether the light is functioning or not, you really need to know the answer, but it seems hard to be confident. And you probably want to err on the side of the most common situation, which is that the lights are working.
> miss the time when "confused" for a computer program was meant in a humorous way
Not sure what about this isn’t funny. Nobody died. And the notion that traffic lights going down would not have otherwise caused congestion seems silly.
> what about the emergency services not being able to reach their destinations?
Did they have documented problems?
This is akin to the Waymos honking at each other at 3AM. Annoying. Potentially dangerous in various circumstances. But ultimately just destructive in a way unlikely to repeat.
Have you seen how human drivers deal with traffic lights and emergency vehicles at the same time? Waymo made the right call to suspend service, they will probably update their playbook to suspend service during power outages in the future.
Humans certainly are imperfect and make mistakes, but will iterate with the understanding that doing nothing at all and blocking emergency vehicles is untenable.
At the least we will fall back to incentive/disincentive social behavior. People will supply ample friendly and unfriendly advice to try to unwind the knot.
Waymo should lose their operating license based on this experience. It's self-evidently dangerous to everyone to be incapable of basic iteration. There's a whole set of law driver's are supposed to follow for handling failed traffic lights. Why have lower expectations of an anonymous car than a human?
> Waymo should lose their operating license based on this experience.
Then everyone should lose their licenses as well by your draconian reasoning. Because…
> There's a whole set of law driver's are supposed to follow for handling failed traffic lights.
And they don’t, it’s chaos.
> Why have lower expectations of an anonymous car than a human?
You obviously have higher expectations for autonomous cars than humans, it is not the other way around for those of us who disagree with you. The only difference is that Waymo can get better with experience and humans generally don’t.
> > There's a whole set of law driver's are supposed to follow for handling failed traffic lights.
> And they don’t, it’s chaos.
Do you live in areas where traffic lights go out regularly?
Because for human driver it is a non-issue. It becomes an all-way stop and you take turns, it is easy. Traffic throughput slows down a bit, but nothing approaching chaos about it. If waymo can't deal with this, that's a problem.
Developing new technologies has risks. In the absence of anything really bad actually happening, I think we can solve the problem by adding new requirements to Waymo's operating license (and all self driving cars) rather than kneecapping the technology.
That sounds plausible. Humans for the most part can usually navigate that situation to a point. It wouldn't surprise me if Waymo cars weren't even trained for this scenario.
I lived in the training zone for both Waymo and Cruise. They were there for literally years before they were offering rides to anyone. The idea that they could train them for emergency scenarios, especially ones that happen so infrequently like a power outage on a route they regularly drive, seems borderline nonsensical, but I honestly don't know if there is a plausible way to do it.
The one time I saw traffic lights go down, it was total chaos. There were two separate crashes that had already happened when I got there, and there would probably be >1 wreck per few minutes with the driving I observed.
I moved from South Africa to Ireland 2 years ago. It was very noticeable to me how drivers in Ireland have no idea what to do when the lights are out. Absolute chaos!
In south africa, traffic lights not working is a daily occurrence. And we've all learned how to navigate a dead intersection wit zero casualties.
Massive 6 way intersections with 2-4 lanes per direction worked perfectly with everyone taking turns to go.
Interestingly in one of the videos online there are several (five I think) Waymos blocking the right two lanes entering the intersection (along with a few others around the other parts of the intersection). While its hazards are still blinking, one of these vehicles moves forward (admittedly just a few feet).
Is this a violation of the California Vehicle Code? Generally it seems to disallow non-emergency vehicles from traveling with blinking lights except for turn signals (and brake lights responding to a braking action).
I couldn't find anything other than their first responders page but IMO any robo taxi operating in a metropolitan area should be publishing their disaster response & recovery plans publicly.
I live in SF in an area that was affected by the blackout. I saw four different Waymos stopped. Three were in the middle of the street. One was along the curb.
My personal opinion. With number of cars I saw flying through blacked out intersection -- major intersections -- I'm very happy that Waymo had a fail safe protocol for such a "white swan"-style event (that is extremely rare, but known-to-happen event).
I saw a damn Muni bus blow through an minor intersection, and was just shaking my head. So many dumbasses behind the wheel, it's miracle no one was killed, and everyone seems to be concerned with "the flow of traffic."
Pretty much no aspect of this event was "extremely rare". PG&E spends the whole year smoking $100 bills and laughing their assess off, then as soon as it rains even a little bit their junk explodes and they pretend they could not have foreseen water existing on Earth. This is not even the first, or second time that this specific substation has burned in living memory. It already burned in 1996 and 2003.
I'm not going to have a semantic argument with you, but I'd consider anything rarer than "a few hours per year" as an "extremely rare event" for the purposes of training autonomous vehicles.
OK well let's argue about the semantics of "autonomous" instead. To me, it means the vehicle's on-board systems should generalize to safe and non-disruptive behaviors under all circumstances. In this instance they should have been able to either navigate to a depot or at least pull off the road.
Everyone should have understood that driving requires improvisation in the face of uncommon but inevitable bespoke challenges that this generation of AI is not suited for. Either because it's common sense or because so many people have been shouting it for so long.
In many versions of road rules (I don't know about California), having four vehicles stopped at an intersection without one of the four lanes having priority creates a dining philosophers deadlock, where all four vehicles are giving way to others.
Human drivers can use hand signals to resolve it, but self-driven vehicles may struggle, especially if all four lanes happens to have a self-driven vehicle arrive. Potentially if all vehicles are coordinated by the same company, they can centrally coordinate out-of-band to avoid the deadlock. It becomes even more complex if there are a mix of cars coordinated by different companies.
I'd say driving only requires not to handle uncommon situation dangerously. And stopping when you can't handle something fits my criteria.
Also I'm not sure it's entirely AI's fault. What do you do when you realistically have to break some rules? Like here, I assume you'd have to cut someone off if you don't want to wait forever. Who's gonna build a car that breaks rules sometimes, and what regulator will approve it?
If you are driving a car on a public street and your solution to getting confused is stopping your car in the middle of the road wherever this confusion happens to arise, and sitting there for however long you are confused, you should not be driving a car in the first place. That includes AI cars.
To be fair 'common sense' and 'many people have been shouting it' about technical matters have a long history of being hilariously wrong. Like claims that trains would cause organ damage to their riders from going at the blistering speed of either 35 or 50 mph, IIRC. Or about manned flight being impossible. Common sense would tell you that launching a bunch of broadcasting precise clocks into orbit wouldn't be usable to determine the distance, and yet here we are with GPS.
Even among people who mean "common" as in "frequent", they aren't necessarily talking about the same frequency. That's why online communication is tricky!
Waymo's performance in this outage was horrible. 6 hours into the blackout there were still many intersections where a Waymo was blocking traffic, unable to navigate out of the way. This should never happen again.
This was very annoying, and made things feel unsafe. Having vehicles stopped blocking visibility when there is no light. Its bad enough we tolerate them stopping and waiting for a pickup and blocking lanes under normal conditions. I had a hard time seeing if there are pedestrians when they’re literally in the cross walk stopped.
I don't understand why everyone is talking about the cars when the bigger issue is why the critical infrastructure (lights) don't have batteries for backup.
Obvious failure should be obvious. Get out of the streets. What happens when one gets a flat tire? Surely it doesn't just stop in the middle of the street, right??
I saw plenty of Waymos managing to make it through intersections. They were slow and tentative, but definitely made forward progress.
I think the emergency "phone home" protocol requires a phone, presumably with enough channel capacity for reasonable video feeds. I wouldn't be surprised if the dead in the road Waymos were lacking connectivity.
There is of course also a possibility that the total demand exceeded the number of people at Waymos available for human intervention.
I think it’s clear that both use cases are a must have during an emergency. Even more, rescue services and stranded people would need all the bandwidth and reception they can get, Waymos shouldn’t be online during such times at all.
FirstNet is AT&T. Verizon and T-Mobile have their own alternatives (FrontLine and T-Priority, respectively). QCI is a fun rabbit hole, because some of these networks share the first responder QCIs with certain business use cases.
I was in the affected area and we effectively lost all but messaging. Not the whole time, but definitely while I was ordering takeout at a place with power. I couldn't get an image to send to a friend.
I lost cell during the whole outage on Verizon, came back immediately when power was restored. There seemed to be some towers up, if i walked down the street I could find one, but plenty were down.
It was predicted by many, including me. It'll be a lot worse in an earthquake where power and cell service are out and there's debris and road damage. Good luck to our first responders.
Coincidentally we were on the Robotaxi during the black out (didn’t know about it, we were going to Japan town from the Mission). Noticed that it navigated through the non-working traffic lights fine, treated it like a stop sign junction. One advantage of building unsupervised system from public version that had to deal with these edge cases all around the country.
Though the safety driver disengaged twice to let emergency vehicles pass safely.
I'd default to assuming it's the respective roadmaps for Waymo and Tesla differed on which things to implement when, not training data, that results in the two behaving different.
50/50 bet it would either go right through or treat it as a stop.
Don’t think I have had a totally inactive light. I have had the power is out but emergency battery turned to blinking red light, and it correctly treats as a stop sign.
FSD is level 2. Level 3 doesn't require the human driver to monitor the outside environment, only take over when requested. Tesla also doesn't report data from FSD under L3 reporting requirements anywhere in the US.
As a society, as a whole society, opposed to our narrow interests and point of views as typical affluent HN dwellers, do we really NEED this kind of shit?
SF Muni & BART both stopped service in many areas. Though most of the trains still had electricity, many sensors and control systems were inoperable. Also underground stations had no lighting, so it would be hazardous to allow people to board or exit there.
Waymo's problem is obvious in hindsight, and quite embarrassing for them, but it can be solved with software improvements. Tesla's FSD already treats dark traffic lights as stop signs, so I would bet on Waymo fixing this as soon as they can.
But transportation that depends on infrastructure along the whole route (such as trains and busses powered by overhead lines) are always going to fail in these situations. I think that's acceptable considering how rare these events are.
I don't understand how these cars keep getting stalled for half an hour or something. Surely there must be a team of teleoperators ready to jump in at any time?
"the infra" is cell phone data coverage though. Which was probably congested by all the city residents using theirs instead of their wifi which was down. Would be fascinating to see just how much Internet traffic flows changed during the outage.
Autonomous vehicles should be on the road iff they reduce overall incidents/deaths. Failure to deal with an out-of-distribution scenario would count against this, but may be rare enough to not significantly affect the average.
Read more carefully, Waymo didn’t boast about their numbers and that part of the story was unrelated to the issue except as a measure of impact their shutdown could have.
The word carrying water here is "need." My partner and I ran an errand in the blackout, we didn't need to do it, but we wanted to. I drove us almost exclusively through an area of the city that had power, but we still ran into issues, including multiple disabled Waymos during that trip. Even in the areas with power, people were driving weirdly. If were able to do it over, I would have stayed home.
I was driving across the east side of SF and hit a patch of lights that were out.
The Waymo's were just going really slow through the intersection. It seemed that the "light is out means 4-way stop" dynamic caused them to go into ultra-timid mode. And of course the human drivers did the typical slow and roll, with decent interleaving.
The result was that each Waymo took about 4x as long to get through the intersections. I saw one Waymo get bluffed out of its driving slot by cross traffic for perhaps 8 slots.
This was coupled with the fact that the Waymos seemed to all be following the same route. I saw a line of about a dozen trying to turn left, which is the trickiest thing to navigate.
And of course I saw one driver get pissed off and drive around a Waymo that was advancing slowly, with the predictable result that the Waymo stopped and lost three more slots through the intersection.
On normal days, Waymos are much better at the 4-way stops than they used to be a few years back, by which I mean they are no longer dangerously timid. The Zoox (Amazon) cars are more like the Waymos used to be.
I expect there will be some software tweaks that will improve this situation, both routing around self-induced congestion and reading and crossing streets with dead lights.
Note that I didn't see any actually dead Waymos as others have reported here. I believe this is an extreme failsafe mode, and perhaps related to just too much weirdness for the software to handle.
It would be interesting to see the internal post mortem.
Failed pretty badly but no reported injuries or even accidents so not that badly.
And if you’re Waymo, it’s a short-term reputation hit but great experience to learn from and improve.
>to learn from and improve.
Okay, let's see if they actually do it this time.
I’m curious what’s the regulation in this scenario? In Canada I think light off means 4-way stop signs so everyone obeys that, or at least most of everyone. What’s the situation in SF?
Yes, that is the same law in California, but so many people drift through stop signs that the guidance is close to meaningless.
In addition, there are 4-way stop signs all over SF and tourists regularly comment on how they work here.
The law is clear - yield to the right, but that is a pretty slow system in congested roads.
The local custom in SF is that someone is usually obviously first, rightmost, or just most aggressive, and opposing pairs of cars go simultaneously, while being wary about left turns.
Of course pedestrians have right of way in California, so someone in a crosswalk gives implied right of way to the road parallel to the person's crosswalk.
The result is 2x or better throughput, and lots of confused tourists.
So ... with the lights out on a Saturday before Xmas, there was a mess of SF local driving protocol, irritated shoppers, people coming to SF for Xmas parties, and just normal Saturday car and foot traffic.
I thought Waymo did pretty well, but as I said, I didn't see any ones that were dead in the middle of the street..
Is this not how four way stops work everywhere? I live in Kansas and have previously lived in Chicago, and I feel like both places follow this custom. Only thing that’s different is the laws are followed slightly more rigorously in low traffic areas, but the customary rules are definitely still in play.
Your local custom seems to describe 4 way intersections everywhere.
Thanks, I get the situation is pretty chaos...
SF drivers mostly don’t stop at stop signs. Many do not stop at red lights either.
Legally in the United States a completely dead traffic signal becomes an all way stop.
Thanks you!
I got stuck behind a Zoox in SF trying to cross the street from an alley. There was an endless stream of stop & go traffic and the Zoox refused to push itself into traffic, despite other cars deliberately giving it space. I wasn't sure if honking at it would help or hurt the situation.
From John Ripley on Mastodon:
“Thought of the day, and I wish there were a way to get this to legislators:
Come the next Big One earthquake, all of San Francisco’s emergency services will be blocked by Waymos.”
I’m AMAZED they’re not designed to handle this better. This does indeed seem like a massive problem. “Oops we give up” right when things get the worst? How is this OK?
I’ve been very impressed by Waymo’s more cautious approach. Perhaps they haven’t fully thought through the ramifications of it though.
https://mastodon.social/@jripley/115758725115731454
> Perhaps they haven’t fully thought through the ramifications of it though.
There is a chorus of voices here on HN that have tried to do this openly, obviously, myself included. It seems to be almost universally derided by people who apparently assume that we're just trying to hurt a start up out of anti-environmental sentiment and jealousy.
There are more ways to get "self-driving cars" wrong than there are to get it right. Driving is far more complex than the hackers here on Hacker News seem to want to concede, and even if that wasn't the case, I'm not sure where the sentiment that a trillion dollar corporation is naturally going to implement this system with the best interests of society in hand comes from.
It's a genuine frustration here.
What was the better solution here then? Assuming there's hundreds or thousands of self-driving cars suddenly driving in environment without any traffic lights. In the pictures you can see six Waymo cars at a single intersection. Assuming some of them had passengers should they all try to turn at the intersection anyway, when their LIDAR says the lane is likely free and pull over to the side? Is that the safest option? Should there be human police to direct the self driving cars through intersections? Or wait out the temporary electricity failure?
I believe the answer is far more complicated than it seems and in practice having the cars stay still might have been the safest option any of the parties could agree on (Waymo's office, the city traffic people, state regulators, etc).
There are people thinking this stuff out and those cars can 100% pull over automatically but an explicit choice was made not to do so for safety.
I think part of the problem is they’ve made it our problem.
Look I like Waymo. I think they’re neat and I trust them far more than any of the other companies. But in my mind being able to handle stuff like this is just a requirement to be on the roads in any non-trivial number. Like if they had two vehicles in this happened then OK that’s a problem but it was two vehicles in an entire city.
When you have enough on the road that you can randomly have six at one intersection you should absolutely be able to handle this by then.
I want them to do good. I want them to succeed. But just like airliners this is the kind of thing where people’s safety comes first.
What we saw happen looks like the safety of the Waymo and its passengers came above everyone else despite having no need to do that. There are certainly some situations where just staying put is the best decision.
The power went out and there are no other hazards on the road is not one of them. They made things worse for everyone else on average in a foreseeable situation where it was totally unnecessary. And that’s not OK with me.
This feels like the kind of thing that absolutely should’ve been tested extremely well by now. Before they were allowed to drive in large volumes.
Effectively they’ve turned any edge case into a potential city-wide problem and PR nightmare.
One driver doesn’t know how to handle a power outage? It’s not news. Hundreds of automated vehicles all experience the same failure? National news.
I live in the affected neighborhood. There were hundreds of drivers that did not know how to handle a power outage... it was a minority of drivers, but it was a nontrivial, but nominally large number. I even saw a Muni bus blow through a blacked out intersection. The difference is the Waymos failed in a way that prevented potential injury, whereas the humans who failed, all fail in a way that would create potential injury.
I wish the Waymos handled it better, yes, but I think that the failure state they took is preferable to the alternative.
Locking down the roads creates a lot of potential injuries too.
And "don't blow through an intersection with dead lights" is super easy to program. That's not enough for me to forgive them of all that much misbehavior.
> is super easy to program
What?!? We’re talking about autonomous vehicles here.
Specifically identifying road signs, traffic lights, and dead traffic lights is a narrow problem that has feasible solutions. To the point where we can reasonably say “yeah, this sub-component basically works perfectly.”
Compared to the overall self-driving problem which is very much not a super easy problem.
Right. You know there are humans somewhere in the city who got confused or scared and mess up too. Maybe a young driver who is barely confident in the first place on a temporary permit, or just someone who doesn’t remember what you do and was already over-stressed.
Whatever, it happens.
This was a (totally unintentional) coordinated screw up causing problems all over as opposed to one small spot.
The scale makes all the difference.
Yeah, the correlated risk with AVs is a pretty serious concern. And not just in emergencies where they can easily DDOS the roads, but even things like widespread weaknesses or edge cases in their perception models can cause really weird and disturbing outcomes.
Imagine a model that works real well for detecting cars and adults but routinely misses children; you could end up with cars that are 1/10th as deadly to adults but 2x as deadly to children. Yes, in this hypothetical it saves lives overall, but is it actually a societal good? In some ways yes, in some ways it should never be allowed on any roads at all. It’s one of the reasons aggregated metrics on safety are so important to scrutinize.
We already have a solution, it's written down in the traffic laws. If the signals fail, treat the intersection roughly like a four-way stop. Everybody learns this in drivers' ed. It's not obscure. If the cars can't follow traffic rules maybe they're not ready to be on the streets unsupervised.
That may be the rules for humans, particuarly people who are always in a rush and won't stay still anyway. With a major intersection turned four-way stop you have lots of humans making very complex decisions and taking a lot of personal risk. If multiple self driving cars make the choice at the wrong time you could jam up an intersection and create a worse traffic issue, or kill a passenger.
It's all a careful risk calculation, those self driving cars need to determine if it's safe to continue through an intersection without the traffic lights their computers spent millions of hours to train on (likewise with humans). That's a tough choice for a highly regulated/insured company running thousands of cars.
If anything, their programming should only take such a risk to move out of the way for a fire truck/ambulance.
> If multiple self driving cars make the choice at the wrong time
Would would they do that? It's a hive, isn't it?
The problem seems to be that the Waymo cars did exactly as you requested and treated the intersections like 4 way stops but kept getting displaced by more aggressive drivers who simply slowed and rolled.
How many non-Waymo accidents happened at intersections during this time? I suspect more than zero given my experiences with other drivers when traffic lights go off. Apparently, Waymo's numbers are zero so humans are gonna lose this one.
The problem here is that safety and throughput are at odds. Waymo chose safety while most drivers chose throughput. Had Waymo been more aggressive and gotten into an accident because it wouldn't give way, we'd have headlines about that, too.
The biggest obstacle to self-driving is the fact that a lot of driving consists of knowing when to break the law.
> The problem here is that safety and throughput are at odds. Waymo chose safety while most drivers chose throughput.
Did they? They chose their safety. I suspect the net effect of their behavior made the safety of everyone worse.
They did such a bad job of handling it people had to go around them, making things less safe.
We know what people are like. Not everyone is OK doing 2-3 mph for extended time waiting for a Waymo to feel “safe”.
Operating in a way that causes large numbers of other drivers to feel the need to bypass you is fundamentally worse.
> Did they? They chose their safety. I suspect the net effect of their behavior made the safety of everyone worse.
There is no viable choice other than prioritizing the safety of your rider. Anything less would be grounds for both lawsuits and reputational death.
The fact that everybody else chose throughput over safety is not the fault of Waymo.
Will you also complain when enough Waymo cars start running on the freeways that a couple of them in a row can effectively enforce following distances and speed limits, for example?
Obstructing traffic is also against the law.
Something I had pounded into me when I drove too slowly and cautiously during my first driving test, and failed.
Those Waymos weren't moving which is a pretty egregious example of obstructing traffic.
An old rule of thumb is every time a service expands by an order of magnitude there are new problems to solve. I suspect and hope this is just Waymo getting to one of those points with new problems to solve, and they will find a way to more graciously handle this in the future.
> Will you also complain when enough Waymo cars start running on the freeways that a couple of them in a row can effectively enforce following distances and speed limits, for example?
In my state, that would itself be a traffic violation, so yes I would. The leftmost lane on an interstate highway is reserved for passing. An autonomous vehicle cruising in that lane (regardless of speed) would therefore be programmed in a way that deliberately violates this law.
Enforcement is its own challenge, whether robots or humans.
> The leftmost lane on an interstate highway is reserved for passing.
Sadly, in most states, this is not true anymore. Most of those laws have been repealed.
I was very pleasantly surprised when I was in Colorado that they had explicit signs saying that if you had 5 (I think) or more cars behind you that you were supposed to pull right and let them pass.
However, I wasn't really thinking about a Waymo cruising in the left lane but simply 4 or 5 Waymo's in the right lane going right at the speed limit with proper following distance. That's going to effectively lock the right lane to the speed limit which then means that even a single other car would lock the left lane to the speed limit as well. Basically, even a couple of Waymos in the right lane would drop freeway speeds dramatically.
> Assuming there's hundreds or thousands of self-driving cars suddenly driving in environment without any traffic lights.
Self-driving cars should (1) know how to handle stops, and (2) know that the rules for a failed traffic light (or one flashing red) are those for an all-way stop.
> What was the better solution here then?
Just pulling over and getting out of the way really would help. There's no reason a human couldn't do the same safely. Not beta testing your cards on public roads would really be ideal. Especially without human drivers ready to take over.
Tbh I'm surprised waymo didn't have remote monitors who could handle cars at intersections or safely pull to the side
Uh, how about having their remote driver staff take over?
> but an explicit choice was made not to do so for safety.
You know this how?
That’s what they usually do. The assumption here is that due to the blackout or some other related issue the human drivers were unavailable.
However even if that’s not true if they have more cars than human drivers there’s gonna be a problem until they work through the queue. And the bigger that ratio, the longer it will take.
I guess that in a blackout they should just have the cars park somewhere safely. Maybe it'd be best to never have more cars on the road than assisting/available human drivers. As soon as no human drivers are available to take over for outage/staffing/whatever reason all cars should just pull over and stop.
This only works if they have cell service and enough human drivers to handle all of their cars.
The better solution? To not fetishize technology.
> There is a chorus of voices here on HN that have tried to do this openly, obviously, myself included.
Maybe I'm reading things wrong, but it sounds like the top comment wants waymo to be better, and you want waymo to be off the roads. You're not talking about the same kind of "thinking through the ramifications".
> I'm not sure where the sentiment that a trillion dollar corporation is naturally going to implement this system with the best interests of society in hand comes from
The sentiment comes from the corporation itself. With this much money at stake you know they have a hand in steering the conversation and that includes on sites such as this.
Willfulness.
Waymo may discover that heavy equipment (large fire trucks can easily push Waymo out of the way if it can find somewhere to push it to) WILL move the cars (at least if there is no one in them at the time) in such cases. I recall the scenes during recent wildfires where abandoned cars were blocking roads and a skip loader was just picking up the cars and dumping/pushing them to the side of the road/over the edge - causing extensive damage to some of them.
Decades ago I recall talking to a fireman expressing a question of what happened if there was a car blocking their access in an emergency and he made it clear that the bumper on the front of the truck and the truck's healthy diesel engine would usually take care of the problem very quickly.
There are people in the Waymos. It's like not moving parked cars.
The jaws of life can make quick work of any locked doors before continuing to bulldoze a car out of the way.
I wouldn't cry for waymo if a bunch of their cars got bulldozed out of the way but that's still unacceptable since it slows down emergency vehicles and first responders.
>I’m AMAZED they’re not designed to handle this better.
This has been the MO for "tech companies" for the past 20 years. Meanwhile I'm told I'm paranoid when the industry of "move fast or break things" decides to move into mission/safety critical industries and use its massive wealth to lobby for deregulation to maintain its habits.
We certainly have BS regulations done to constrain competition. But I'd wager a good 80% of them exist for good reason.
In cases where the traffic signal is not working, it is known that the FSD has to take on a more challenging role of reading traffic agent gestures. I think they have that functionality built in. But not when neither traffic signal is working nor traffic agent is present.
The basic thing is to treat everything like a four-way stop sign.
I would just push them all out of the way with my fire truck, I mean one fire truck could probably clear 6-8 Waymos at a time, right?
Fire trucks are very expensive to be playing bumper cars with
Modern fire trucks, and police cars usually, are built to be able to push vehicles out of the way. It's a very common need.
(Not an argument against Waymo doing better in this situation though!)
If we reach the point of needing to forcicbly move mass numbers of cars off the road for fire trucks, that's a dire situation where routine cost/benefit analysis has already gone out the window.
if google's property is blocking the road, google can pay for the damage
Just paying for the physical damage isn't enough. It should also come with massive fines for obstructing emergency vehicles.
That doesn't bring people back to life or restore quality of life for life-changing injuries.
As long as you're happy that if your property ever blocks the road, you will pay for the damage too.
> As long as you're happy that if your property ever blocks the road, you will pay for the damage too.
Pretty sure that's always the expectation? It's typical to tow illegally parked cars, smash windows to run hoses through cars blocking hydrants, etc.
The only unusual thing here would be holding a corporation to account the way we hold individuals to account.
No I won't be happy about it, but yes if I block emergency services and they need to damage my property then I am absolutely the one who should have to pay.
absolutely
don't be a dick, don't block the road
Duh?
They have massive steel bumpers, pushing them away slowly seems mostly harmless (not at maximum velocity lol)
Although it would be amusing for them to do it at high velocity if the cars (and surrounding cars if any) were "dead heading" or had no humans in them for other reasons (perhaps because the humans had fled the vehicles upon seeing the fire truck headed their way!).
True, but in a 'Big One' event i dont think we would care.
You care because it takes valuable time to move things out of the way and in an emergency time is often the one resource you don’t have.
In an emergency, if the only options are taking time moving things out of the way or not being able to move, you move things out of the way.
GP’s point is that’s a false dichotomy.
Yes that’s the correct decision when those are the only options, like if a car has stalled or the driver just got out and ran away.
In this case there’s a third option: the computer that’s still perfectly functional should have been able to get out of the way on its own. And legally all drivers are required to.
I assume that applies to robots as well, if it doesn’t it absolutely should.
Surely if the Big One hits then all of the metropolitan areas on the West Coast would be gridlocked in scenes reminiscent of zombie apocalypse movies anyway? I guess we won't know until it happens for sure, but I can't imagine it would be easy for emergency services to get around with or without Waymo.
It’s not gonna be good. But you want it to be a gridlock because the cars can’t get out fast enough because there’s too many cars on the road.
Not because a bunch of cars that are perfectly capable of moving are just sitting there blocking things purposefully waiting for the driver in the sky to take over.
And what if, due to $BIG_DISASTER they won’t be able to for a week?
In case of a natural disaster, it’s guaranteed that human drivers will abandon their cars on the road and cause gridlock. It happens all the time. Emergency vehicles are built to handle it.
During Japan's 2011 earthquake, many roads were gridlocked by human drivers.
> Come the next Big One earthquake, all of San Francisco’s emergency services will be blocked by Waymos
Were any emergency vehicles actually blocked?
We have an actual failure here–step one is identifying actual failures so we can distinguish what really happened from what hypothetically could.
I don’t know. But if human drivers are having to go around them, they’re not doing the right thing.
They need to drive or pull over. Never just stop there in the road and wait.
> if human drivers are having to go around them, they’re not doing the right thing
They're not. But it's also not a disaster. Pretending it is on Twitter is pandering, not policymaking.
> They need to drive or pull over. Never just stop there in the road and wait
Agreed. Waymo has a lesson to learn from. Sacramento, and the NHTSA, similarly, need to draw up emergency minimums for self-driving cars.
There are productive responses to this episode. None of them involve flipping out on X.
> But it's also not a disaster
Because it’s a power outage. If we instead learned about this during a real disaster people could have died because these things were let on the road without planning what they should do in abnormal circumstances.
We’re lucky it’s not a disaster.
> If we instead learned about this during a real disaster people could have died
This is universally true. The question is how bad could it have been, and in which cases would it have been the worst?
> We’re lucky it’s not a disaster
This is always true. Again, the question is how lucky?
We have an opportunity to count blocked emergency vehicles and calculate a hypothetical body count. This lets us characterize the damage. But it also permits constraining hysteria.
> We’re lucky it’s not a disaster.
I'm sure that if this was something predictable like a cyclone or wildfire, Waymo would still have 100% of their nightly traffic on the road, right? And SFFD would not be able to do what they normally do when they can't get support, which is hop into the car and use the controls to manually move it?
Or... maybe Waymo HAS considered what their cars should do in abnormal circumstances and this kind of outcome was considered acceptable for the number of cars and the nature of the "disaster"?
> Waymo has a lesson to learn from.
At what point can we be spared from having Waymos lessons inflicted upon us
> At what point can we be spared from having Waymos lessons inflicted upon us
Again, we had a real event happen. Not hypothetical. What was the actual cost inflicted?
In this country, if heart disease or cancer doesn't kill you, a car probably did.
Until "Waymos lessons" are killing people at that rate, I am 100% OK with a Waymo making my trips an extra 5 minutes longer every 50th trip or whatever else the real stat is.
I was curious if Waymo has even been involved with a crash that killed someone, so I looked it up. The answer is yes - there was a Tesla going 98mph in SoMa whose driver died after hitting a Waymo. Clearly the Waymo's fault!
When humans can cause fewer accidents and fatalities than Waymo on average. People are still inflicting those lessons on us.
When we learn our lesson that letting companies beta test on public roads consequence free is just another cost to the rest of us so that a small number of people can enrich themselves at our expense.
Whenever they become so much a problem that they counterbalance public and private interests in having and improving robotaxis. For most people, we are nowhere near that.
No, it's not a disaster, but with a little imagination it could be a hormetic innoculation.
> They're not.
They are. I did myself yesterday because one was sitting at the front of a turning lane at a dead light, just waiting there forever with the blinker on.
waymos shouldn't exist, and san francisco shouldn't just be a experimentation lab for tech companies
A fire truck can simply push the waymo out of the way.
It can’t push a block of gridlocked traffic that cannot move because of the dead waymos present out of the way.
Unless a few Waymos have gridlocked traffic, I'm not sure you can still blame them for this.
It can do that with a normal driver too. Doesn’t make it ok for there to ever be a situation where they need to when the target vehicle/driver is just fine and capable of doing it themselves.
> A fire truck can simply push the waymo out of the way
Sure, but it would be notable if one had to. If none had to, we have a problem to solve, not a catastrophe.
I'm surprised the don't know to treat it as a 4-way stop, either. This kind of outage is pretty common in Phoenix, too, which is another major Waymo market. It probably happens to at least some part of the city every monsoon season.
What if there was a herd of people off-shore on-call willing to basically "RDP in" and take over control (human takeover) of the entire fleet when needed? I could see that being an attractive pitch.
Latency makes this hard even with local connections, it’s essentially impossible due to physics to do it offshore.
And I believe Waymo remote access only allows providing high level instructions (like pull over, take the next right, go around this car, etc) precisely because full direct control with a highly and variably latent system is very hard/dangerous.
And in an emergency situation you’re likely to have terrible connectivity AND high level commands are unlikely to be sufficient for the complexity of the situation.
I suspect that in a large scale disaster/emergency the communications systems may be disrupted and it may not be possible to remotely control the vehicles.
Perhaps in such cases they can pull over in a safe place, or if they have an occupant ask them if they wish to continue the journey or stop.
Perhaps they already do this, I have no experience with autonomous vehicles.
This very scenario could involve no cell signal.
Waymo will get better at this.
But even without them getting better, as far as I know there were zero waymo fatalities due to this.
That's more than I can say about Helene, where there was at least one fatality due to traffic light outages.
Lets not forget that a big part of why we want Waymo is that it has already lead to a dramatic decrease in fatal accidents. They are a great company that will do a lot of good for the world. One bad night (in which noone was hurt, in part because of their cautiuosness) shouldn't negate that.
"Hey guys what are you complaining about? We didn't (directly) kill anyone!"
Seeing as how literally nobody died, I'm not sure if I agree with your sentiment.
I was curious if Waymo has even been involved with a crash that killed someone, so I looked it up. The answer is yes - there was a Tesla going 98mph in SoMa whose driver died after hitting a Waymo. Clearly we should shut down Waymo until they can handle that situation!
You're being sarcastic, but it's a valid point. I'd love to know if there were any traffic fatalities at all during the affected period. Chances are there were and that they were due to human error.
In my experience, humans respond incredibly poorly to traffic lights being out. There's no sense or reason, just people deciding to drive across the intersection when they feel like it's okay.
Presumably Waymo will make sure they can handle this situation in the future, but I'm not sure there's a really satisfactory solution. The way you're supposed to handle an intersection with no lights (treat it as a stop sign intersection) doesn't work very well when no one else is behaving that way.
That wasn’t my experience, having just driven across the city and back during tonight’s outage. It was actually weirdly inspiring how well people coordinated at so many of the powerless intersections.
There was a lot of confusion, and some people took advantage of it to rush through, but for the most part it was pretty orderly. Which makes sense because in many parts of the world where there are no traffic lights or stop signs, people get on just fine.
The Waymo’s, on the other hand, were dropping like flies. While walking from Lower to Upper Haight I spotted a broken Waymo every handful of blocks. The corner of Haight & Fillmore was particularly bad, with 3 of them blocking traffic in both directions — in the path of both the 7 and 22 bus lines.
There have been some experiments that suggest that traffic flow is both safer and more efficient if you just turn disable the signals entirely. I doubt it applies to all, or even most, situations but it's definitely food for thought: https://www.npr.org/2008/01/19/18217318/german-towns-traffic...
The general wisdom is to treat a broken light like a stop sign. So it really comes down to how well that wisdom is absorbed when the time comes.
Herd mentality also helps here. We see the first few people do this and we will follow along pretty quickly.
>in many parts of the world where there are no traffic lights or stop signs, people get on just fine
Well, sort of. Road injuries / fatalities in countries without these kinds of regulations are about an 3-4x higher than in those that do have them.
https://www.who.int/publications/i/item/9789241565684
Number of traffic accidents went down by 50% in this town.
https://worksthatwork.com/1/shared-space
I think a significant factor helping that to work is the mixing of all traffic on the street. I've noticed that in LA's Skid Row, where homeless people are constantly moving into the street on foot or on bicycle and they walk around in vehicle lanes pushing shopping cart armadas and so on, drivers are more cautious than usual and I see, if anything, less reckless driving and close calls there than in other parts of downtown, where pedestrians stick to the sidewalk and distracted or car-brained drivers don't look out for them. Just anecdotal observation, of course.
Different things. A country with lax rules is not the same as a specific environment with shared spaces, where according to known data it's safer to eliminate some specific kind of regulation and let the remaining part take over.
It’s not lax rules, in many cases it’s just alternative coordination — eg roundabouts
> with 3 of them blocking traffic in both directions — in the path of both the 7 and 22 bus lines.
wow, cascading failures. I'll bet this is the tip of the iceberg.
i had essentially the opposite experience in close proximity.
I've been through long blackouts.
My own experience has given me a somewhat more-nuanced take.
At first, it's akin to the path of evil. Way too many people just zoom through intersections with dark traffic lights like they're cruising unimpeded down the Interstate, obvious to their surroundings. Some people get grumpy and lay on the horn as if to motivate those in front of them to fly through themselves.
But many people do stop, observe, and proceed when it is both appropriate and safe.
After awhile, it calms down substantially. The local municipality rounds up enough stop signs to plant in the middle of the intersections that people seem to actually be learning what to do (as unlikely as that sounds).
By day 2 or 3, it's still somewhat chaotic -- but it seems "safe" in that the majority of the people understand what to do (it's just stop sign -- it may be a stop sign at an amazingly-complex intersection, but it's still just a stop sign) and the flyers are infrequent-enough to look out for.
By day 5 or 6, traffic flows more-or-less fine and it feels like the traffic lights were never necessary to begin with. People stop. They take turns. They use their turn signals like their lives depend on it. And the flyers apparently have flown off to somewhere else. It seems impossible to behold, but I've seen it.
But SF's outage seems likely to be a lot shorter than that timeline, and I definitely agree with Waymo taking the cautious route.
(but I also see reports that they just left these cars in the middle of the road. That's NFG.)
Huh, here in Germany we have street signs (mostly of a "you are the priority road" 45° rotated square yellow-on-white "sunny side up egg" sign and the "you are not the priority road" down-pointing white-on-red triangle; for 3-way if the priority road isn't the straight road or the concept of straight is ambiguous, there's a supplemental sign depicting the path of the priority road) permanently on traffic lights; it's also common enough for non-major roads to have the lights turned off at night so drivers tend to be familiar with falling back to the signs when the lights are off.
In absence of priority roads there is also the "right before left" rule which means that the car coming from the right if they would conflict in time is the car that has priority. It's also always illegal to enter an intersection if you can't immediately clear it; that seems to work better when there are no green traffic lights to suggest an explicit allowance to drive, though.
Sure, but we're pretty far from Germany over here.
In the States (or at least, every US state that I'm familiar with -- each one is free to make their own traffic rules, similar to how each EU member state also has their own regulatory freedoms), a dark/disabled/non-working traffic light is to be treated as stop sign.
For all drivers, in all directions of travel: It functionally becomes a stop sign.
That doesn't mean that it is the best way, nor does it mean that it is the worst way. It simply is the way that it is.
How does "you can only piss with the cock you've got" translate to German slang?
Treating it like a stop sign also doesn't work very well when there are huge amounts of pedestrians. As a pedestrian that yesterday meant I got the right of way all the time. For cars it was mayhem downtown.
In contrast many years ago I lived at an intersection that had almost no pedestrians back then and a few times for a power outage limited to our building and that intersection. I enjoyed standing on my balcony and watch traffic. It mostly worked well. Cars did treat it like a intersection with stop signs. There two issues happened though. One was when there was no car already stopped and about 10%-20% of drivers didn't realize there was an intersection with lights out and just raced through it. The other ironically were bicyclists. 90% of the just totally ignored there was an intersection. That was especially scary when they arrived at the same time as one of those cars who didn't realize it either.
> Treating it like a stop sign also doesn't work very well when there are huge amounts of pedestrians. As a pedestrian that yesterday meant I got the right of way all the time. For cars it was mayhem downtown.
I’ve long been curious if people in S.F. who are used to Waymo “behavior” – including myself – behave differently when a Waymo is involved. For the most part, Waymos are extremely predictable and if you’re on the road as a pedestrian, cyclist, or car, are you more aggressive and willing to assert your turn in the road? Curious if Waymo has done a psychological study of how we start to think about the vehicles. I know many runners, for example, that’ll stride right into the crosswalk in front of a Waymo but wouldn’t dare to do that in front of other cars. Similarly, anecdotally people treat public transit vehicles differently than a civilian car: folks are more willing to let them drive even if out of natural turn.
I saw this recently when the lights were out at an intersection in Manhattan. People kept on driving and almost hitting pedestrians and cars. I called 911 and then directed traffic for 15 minutes until DoT came out and put up a temporary stop sign.
In Germany most traffic lights have a full set of traffic signs that are in effect in the rare occasion that the light is out.
Same setup in the Netherlands, there are right of way signs everywhere that apply when the lights don't work.
One interesting effect is that there are also often pedestrian crossings that have priority over everyone. Normally those are limited by lights, but without lights a steady stream of pedestrians stops all traffic. Seen that happen in Utrecht near the train station recently, unlimited pedestrians and bikes, so traffic got completely stuck until the police showed up.
Except in NL you guys actually have someone sit and think about the flow of traffic, in most other countries the design is usually pretty much random. In my parents' city in Poland they decided to turn off the traffic lights at night and the result is much higher number of accidents. Funnily, my father failed to yield the right of way exactly when talking about the issue. Yes, theoretically all intersections do have full set of signs, but in practice the visibility of those signs is extremely limited.
> Same setup in the Netherlands, there are right of way signs everywhere that apply when the lights don't work.
Most places in Poland have this exact setup. And I say most because I have not seen one that does not, but I am guessing they exist. Maybe some of them have bad visibility even?
If one does not respect the yield sign that does not seem a signaling problem.
Some don't in Finland. But then it is back to basic rule of yielding to traffic from right. Which is pretty common on smaller roads so drivers should think about it enough.
> that does not seem a signaling problem.
Average signaling in Poland be like
https://files.catbox.moe/ipl96o.png
Are the words critical? As a German I read that as "you are not the priority road; you may only go left or right, but not straight"?
The words sum up to "don't go straight unless you have business being there" but still, it really sums up the approach to signaling. Nobody in the whole command chain sat and thought "that might be a tiny bit difficult to parse".
I mostly read those signs as “don't go there unless you already know you can” which as a “tourist” I just assume can't (unless I'm a local, and figure it out).
What I've recently found troubling is the places that use similar signs for emissions controls. With a rental you usually have a recent enough car that you can ignore those.
Being able to distinguish between “low emissions zone, but any car from this decade can go in” and “local traffic only, you need to live in this neighborhood to enter” in a foreign language, bit me a couple of times while traveling.
san francisco really really needs this.. the traffic lights were incredibly hard to see with the fog and rain..
> In my experience, humans respond incredibly poorly to traffic lights being out.
My purely anectodotal experience is that the response is variable and culturally dependent. Americans tend to treat any intersections with a downed stoplight as a multi-way stop. It's slow but people get through. I've experienced other countries where drivers just proceed into the intersection and honk at each other. (Names withheld to protect the innocent.)
It seems a bit like the Marshmallow test but measures collaboration. [0]
[0] https://en.wikipedia.org/wiki/Stanford_marshmallow_experimen...
Where I'm from (a relatively rural country) they just get treated like give way signs you'd have on a country road, the larger road generlly has priority but as it's not clear they'd be more cautious too.
The problem is when one road is busy enough the other doesn’t get a “look in”.
Rural drivers will know to turn right and make a u-turn, but city drivers may not know that trick.
One could argue that it's "cultural", but California state law says this about the situation:
> Traffic Light Not Working
When a traffic light is not working, stop as if the intersection is controlled by STOP signs in all directions. Then proceed cautiously when it is safe to do so.
https://www.dmv.ca.gov/portal/handbook/california-driver-han...
I thought the traffic went pretty well tonight in San Francisco considering we had this major issue.
I've seen 4-way 1-lane intersections behave well.
But those complex multiple lanes in all directions + turn lanes...
They do break down. I think they are a breeding ground for confusion and frustration.
Not really, they just treat it the intersection as if it were ruled by stop signs. It’s not always easy to keep track of who goes next but overall people can handle themselves pretty well in those cases.
Neither a lack of traffic lights nor cell service should cause the Waymos to stop in the middle of the road, that’s really troubling. I can understand the system deciding to pull over at the first safe opportunity, but outright stopping is ridiculous.
What makes you think it was either?
AIUI, it was the irregularity of the uncontrolled intersections combining with the “novel” (from the POV of the software) driving style of the humans. In dense areas during outages signaled intersections don’t actually degrade to 4 way stops, drivers act pretty poorly.
The normal order and flow of traffic broke down. The software determined it was now outside its safe parameters and halted.
Certainly not ideal, and the should be a very strong regulatory response (the gov should have shut them down), and meaningful financial penalties (at least for repeat incidents).
> The normal order and flow of traffic broke down. The software determined it was now outside its safe parameters and halted.
And my question is why did it halt instead of pull over?
How much to move once you decide things have gone off the rails is a hard question, but I’d assume they have put some thought into it.
Then they shouldn't be on the road. What happens if there's an earthquake? The Waymos are going to be blocking everyone else.
Perhaps this is by design. Cruise had a failsafe system that detected a collision and decided to pull over but by pulling over it dragged a person underneath the car (or something close to this scenario). Maybe this dumb failsafe was designed not to repeat Cruise's mistakes?
Certainly a better way to handle this would have been to pull over. I think stopping where ever it happened to be is only acceptable if the majority of sensors fail for some reason
I was there. I encountered multiple stopped Waymos in the street. It was annoying, but not dangerous. They had their lights on. Any driver following the rules of the road would get around them fine. It was definitely imperfect, but safe. Much safer than the humans blowing through those very same intersections.
When I was a young man, I worked at a restaurant, and the lights went off.
I being the hero I was, wanted to keep the show running, bought some candles, ovens worked fine, water worked fine (for now). I wanted to charge cash. But eventually big boss came and shut us down since light wasn't coming.
And he was right, cooking and working under those conditions is dangerous for the staff, but also for the clients, without light you cannot see the food, cannot inspect its state, whether stale, with visible fungi, etc...
Yes, the perfect worker would still operate under those conditions, but we are not perfect, and admitting that we only can provide 2 or 3 nines, and recognizing where we are in that 0.01% moment, is what keeps us from actually failing so catastrophically that we undo all of the progress and benefits that the last bit of availability would have allowed us.
Waymos rely on remote operators to take over when the vehicle doesn't know what to do, and obviously if the remote connection is gone then this is no longer available, and one might speculate that the cars then "fail safe" by not proceeding if they are in a situation where remote help is called for and inaccessible.
Perhaps traffic lights being out is what caused the cars to stop operating autonomously and try to phone home for help, or perhaps losing the connection home is itself enough to trigger a fail safe shutdown mode ?
It reminds a bit of the recent TeslaBot video, another of their teleoperated stunts, where we see the bot appearing to remove a headset with both hands that it wasn't wearing (but that it's remote operator was), then fall over backwards "dead" as the remote operator evidentially clocked off his shift or went for a bathroom break.
That’s clearly unacceptable. It needs to gracefully handle not having that fallback. That is an incredibly obvious possible failure.
Things go wrong -> get human help
Human not available -> just block the road???
How is there not a very basic “pull over and wait” final fallback.
I can get staying put if the car thinks it hit someone or ran over something. But in a situation like this where the problem is fully external it should fall back to “park myself” mode.
> How is there not a very basic “pull over and wait” final fallback
Barring everything else, the proper failsafe for any vehicle should be to stop moving and tell the humans inside to evacuate. This is true for autonomous vehicles as well as manned ones–if you can't figure out how to pull over during a disaster, ditching is absolutely a valid move.
If the alternative is that the vehicle explodes, sure. And since GP did say "final fallback", I suppose you're right. But if the cars are actually reaching that point, they probably shouldn't be on the road in the first place.
The not-quite-final fallback should be to pull over.
Yeah. I wasn’t considering people, just getting the car out of the way.
I wasn’t considering people taking it as a given that any time the car gives up the doors should be unlocked for passengers to leave if they feel it’s safe.
And as a passenger, I’d feel way safer getting out if it pulled over instead of just stopped in the middle of the street and other cars were trying to drive around it.
No one should ever be trapped inside by the car.
They now apparently run these things on the interstate, the car needs to do more than just stop.
Seems like a power outage is a an obvious use case Waymo should have foreseen.
Makes me think there are likely other obvious use cases they haven’t thought about proactively either.
> Seems like a power outage is a an obvious use case Waymo should have foreseen
We have zero evidence a power outage wasn't foreseen. This looks like a more complex multi-system failure.
Does it matter?
Once you’re on public roads, you need to ALWAYS fail-safe. And that means not blocking the road/intersections when something unexpected happens.
If you can physically get out of the way, you need to. Period.
> Does it matter
Yes. OP is inferring Waymo's internal processes from this meltdown. ("Makes me think there are likely other obvious use cases they haven’t thought about proactively either.")
If Waymo literally didn't foresee a blackout, that's a systemic problem. If, on the other hand, there was some weird power and cellular meltdown that coïncided with something else, that's a fixable edge case.
> > Does it matter
> Yes. OP is inferring Waymo's internal processes from this meltdown. ("Makes me think there are likely other obvious use cases they haven’t thought about proactively either.")
No, I'm not inferring internal processes.
I'm guessing level of critical thinking.
When you are creating autonomous vehicles, one of the things that you want to risk assess and have mitigation for is what you want the vehicles to do in case the systems they depend on fail (e.g. electricity, comms).
Now, it could be that the team has anticipated those things but some other failure in their systems have caused vehicles to stop in the middle of intersections, blocking traffic (as per article).
I'm super curious to learn more about what Waymo encountered and how they plan to up their game.
> I'm not inferring internal processes…I'm guessing level of critical thinking
Genuine question: how do these differ? Isn’t the level of critical thinking of Waymo’s employees internal to it? (What’s the mens rea analogue for a company?)
The "coinciding problems" should be an assumption, not a edge case we reason away. Because black swan events are always going to have cascading issues - a big earthquake means lights out AND cell towers overloaded or out, not to mention debris in streets, etc.
What they need is a "shit is fucked fallback" that cedes control. Maybe there is a special bluetooth command any police or ambulance can send if nearby, like clear the intersection/road.
Or maybe the doors just unlock and any human can physically enter and drive the car up to X distance. To techies and lawyers it may sound impossible, but for normal humans, that certainly sounds better. Like that Mitch Hedberg joke, when an escalator is out of order it becomes stairs. When a Waymo breaks it should become a car.
> Or maybe the doors just unlock and any human can physically enter and drive the car up to X distance.
Do the even have physical controls to do that at this point?
I’ve never been in one so I don’t know how different they are from normal cars today.
The Waymos still have all their normal driver controls. There is a process where law enforcement can enter the vehicle, call Waymo and verify their law enforcement status, and then switch the vehicle into manual mode and drive it as normal.
Here is their instructions for law enforcement in the Waymo Emergency Response Guide:
https://storage.googleapis.com/waymo-uploads/files/first%20r...
Ok. Thanks. I must have been thinking of something else.
Didn’t Google have little self-driving vehicles without controls that were limited to pre-planned routes on non-public roads on their campus?
Obviously a hugely different problem domain.
>If Waymo literally didn't foresee a blackout, that's a systemic problem.
I agree with this bit
> If, on the other hand, there was some weird power and cellular meltdown that coïncided with something else, that's a fixable edge case.
This is what I have a problem with. That’s not an edge case. There will always be a weird thing no one programmed for.
Remember a few years ago when a semi truck overturned somewhere and poured slimy eels all over the highway? No one‘s ever gonna program for that.
It doesn’t matter. There has to be an absolute minimum fail safe that can always work if the car is capable of moving safely. The fact that a human driver couldn’t be reached to press a button to say to execute that is not acceptable. Not having the human available is a totally foreseeable problem. It’s Google. They know networks fail.
This isn't to disagree with your overall point about proper emergency mitigation and having humans available.
> Remember a few years ago when a semi truck overturned somewhere and poured slimy eels all over the highway? No one‘s ever gonna program for that.
While the cause is unusual, this is really just three things that everyone absolutely should be programming into their autonomous vehicles: accidents, road debris, and slick conditions.
Certainly. That one was interesting both because of the odd specifics of it and because it made the road more slippery than any normal accident where just a bunch of boxes of random dry goods fell over.
It just happens to make a fantastic example of “thing no one is ever going to foresee“.
If there wasn’t footage how many people would even believe it happened?
A fail-safe is EXACTLY blocking roads at intersections without power, not proceeding through intersections without power. It's much safer to be stopped than to keep going. I honestly wish the humans driving through blacked out intersections without slowing down in my neighborhood last night actually understood this.
It’s not a fail-safe. It’s a different failure mode. Jamming up traffic, including emergency traffic, creates systemic problems.
It’s a bit like designing an electronic lock that can’t be opened if the power goes out. If your recourse to exiting a dangerous situation becomes breaking the door, then the lock is unsafe.
Fail-safe means "in a situation where the function fails, fail in a way that doesn't cause injury" -> the cars didn't know how to proceed, so they stopped, with their lights on, in a way that any attentive driver could safely navigate... which is a failing safe.
The alternative here, is a protocol that obviously hasn't been tested. How on earth are you going to test a Waymo in blackout conditions? I would rather have them just stop, than hope they navigate those untested conditions with vulnerable pedestrians and vehicles acting unpredictable.
Simulate them on a test course? There are absolutely places with street lights and everything that you could test something. Hell since they don’t need to work you can just have some put up in a parking lot to test with. Who cares.
You don’t need to wait for a city blackout to actually test this kind of scenario.
The thing still has cameras. And LIDAR. It should be fully capable of pulling over on its own safely. Why would not having a traffic light prevent that?
Humans are expected to negotiate this. The robots should be too. That’s part of driving. If the lights fail, the law tells you what you’re supposed to do. And it is not stop the intersection.
> Simulate them on a test course?
Yes, what’s the worst that could happen… oh wait… people literally getting killed.
When did a Waymo kill someone on a closed test course?
I'm suggesting that, perhaps, the vehicle will not preform the same way in a dangerous, real world scenario as it would in a training exercise.
> Fail-safe means "in a situation where the function fails, fail in a way that doesn't cause injury"
In a very local sense, this is true. In terms of the traffic system, this can create a systemic problem if the stoppage causes a traffic jam that creates problems for emergency vehicles.
Thus it is a _different_ failure mode.
If someone stops in the middle of traffic because they’re lost, their GPS went out, or they realized that they’re unsafe to drive, we don’t celebrate that as the driver entering a fail-safe mode. We call that “bad judgment” and give them a ticket.
If it precipitates a larger problem where lives are lost, they may be in considerable legal or financial trouble.
I don’t see why we should treat Waymo any differently.
Traffic doesn’t cause injury. Why are we concerned about traffic flow in a blackout situation. The cars stopped at intersections, EMS could use the oncoming lanes. I’m not seeing how it’s not a fail safe, you’re describing it as not being fail-ideal, and I would agree.
One way roads exist.
I'm confused, is your concern that enough Waymos shut themselves down on a one way road, at the same intersection, so as to block the intersection? Yes, I could see that as being a concern. I suspect it would be reported almost immediately, and would be at the top of the list of for the folks at Waymo to address. The cars weren't abandoned. They eventually moved. Though, I suspect they had to be manually driven (virtually or otherwise), out of the way. I can see how this could be a problem, but considering it would likely be at the top of the list of problems for Waymo -- again, during an emergency -- that I suspect it's not a serious concern in the long run.
Would I preferred that they had a light turn on that was flashing "An unknown emergency is occurring, please park me"? Yes, I think that would be a better solution. I would have preferred better performance from Waymo. My entire point here is that I'm happy that in my neighborhood, the Waymos were acting in a fail-safe manner, rather than just winging it.
An intersection without power is just a 4-way stop.
An intersection without power is supposed to be treated as a 4-way stop. An unfortunately high, nontrivial number of drivers last night were not following that rule.
And yet the humans managed.
Even at a normal four-way stop with stop signs people sometimes blow through it. The Waymo has to handle it.
That’s part of driving.
It can creep through at 3 miles an hour if it thinks that’s what’s safe. All it has to do is get out of the intersection.
The outrage people would rightly have at Waymo allowing a number of its vehicles to blow the lights would be huge. People running blacked out lights is unacceptable.
Who said “blow through”?
Waymos know how to handle 4 way stops.
You're anthropomorphizing. Waymos "know" how to handle the 4-way stops that they've been trained to handle.
> Once you’re on public roads, you need to ALWAYS fail-safe.
Yes.
> And that means not blocking the road/intersections when something unexpected happens.
No. Fail-operational is not the only allowable fail-safe condition for automobiles. For example, it is acceptable for loss of propulsion to cause stop-in-lane — the alternative would be to require high-availability propulsion systems, or to require drivers to always have enough kinetic energy to coast to side. This just isn’t the case.
One can argue that when operating a fleet with correlated failure modes the rules should change a bit, but that’s a separate topic.
Yeah, I was shocked by this. Blackouts in California aren’t some sort of rare event. I’m primed to expect rolling brownouts/blackouts yearly in the summer.
There haven't been rolling blackouts since 2001.
There were significant power shutdowns in California in 2019 (affecting millions of customers in aggregate); the reason for the shutdowns was different from 2001 (preemptive shutdowns when the risk of downed power lines starting wildfires was thought to be high) but the impact on customers is the same: no power for an extended period.
> There haven't been rolling blackouts since 2001.
Waymo operates in more places than the Bay Area. Phoenix, AZ, for example, had widespread power outages in Aug 2025 due to Haboob and Monsoon.
One usage case that I saw myself is when a vehicle is parked such that it will require the other vehicle to go slighty over the curb, in this case the curb is flat so I assuming the parked driver thought it was okay. Every other human driver did okay, but Waymo just refused to put its wheel on the curb and just got stuck. Video here: https://x.com/aaditya_prakash/status/1989444130238259575?s=2...
It also means that their claims of "autonomy" are fraudulent, like most "self driving" cars. A car which depends on powered infrastructure outside the car to drive is not autonomous.
Nearly all humans depend on powered infrastructure outside their car to drive it. You're describing a shortcoming of all 21st century navigation.
I think the 100 or so miles I can generally drive in my car while it still has gas is different from my car just stopping suddenly because of a power outage.
Horse manure. If the power goes out, I can still drive. Including navigating intersections. Same for my cell phone.
Tell that to all of the humans who were capable of driving, but blocked by a fake autonomous car that froze in the middle of the road.
We should put self driving cars on tracks so they are always out of the way and have easily predictable behavior. Maybe we can even link the cars together for efficiency or something like that.
You can further optimize the setup by not installing engines/motors in all of them. So maybe you have one car providing locomotion, with the rest following behind and designed for carrying.
And all the power could just come from a few large centralized facilities that are super efficient. We could just use thin strands of metal to get it to the vehicles over head…
Of course, the maintenance on those wires outside of the city means that you'd make electric trains with large batteries on them instead.
https://evmagazine.com/articles/tesla-launches-first-all-ele...
Expense is correlative to scale, likely it's cheaper to deploy pantographs than battery factories.
Why did India build a high speed freight corridor with overhead power when they could have used batteries instead? Because the quantity of battery to power the trains doesn't exist, and overhead wires do.
I don't think I buy BEV trains to be honest. I'm struggling to think of a proper reason why they might be better compared to normal electric trains.
But the linked article is pretty light on info, so I'll reserve judgement till more info comes to light.
They are good for infrequently used track and places where overhead wires would be in the way, like that very Tesla employee shuttle on it's own track and container ports.
It's not the best way to go for mainline track and not suitable for long distance high speed trains.
If it's infrequently used, a dual mode diesel-electric can fill that use case today.
If we do all of that, then we won't need to train them to know how to operate in traffic. Perhaps we can give them a name in honour of that fact?
Maybe we can do away with the expensive battery if we feed power to these cars using overhead cables.
At this point overhead cables are likely many orders more expensive.
Buses in Paris run with IIRC 60kWh battery and pantograph charger at every other station. Packs (not cells) recently dropped to below $100 kWh. At $6k thats probably what city pays for couple of replacement seats (gold plating et all).
In the original Sim City, having your city entirely use tracks instead of roads was superior because there was no traffic congestion from tracks
And less pollution from cars.
If you're happy to put those tracks on every road, sure. I wonder why no-one's bothered with that before.
Sorry if you're playing in to the joke, I can't tell. Streetcars / trams were widely deployed before they were ripped out for the car, driven by lobbyists interested in selling cars. Wondering why no one has bothered with that is starting from a false premise, because people have bothered with that.
I'm well aware that there are multiple options for public transport, but none of them are actually as flexible or as far reaching as cars/taxis. To me this 'but why not trains' on every article about self-driving cars is a tired meme that fails to address that these are not equivalent options. I might as well say "Well, why don't we get rid of these expensive rails and fixed timetables and just lay down some cheap concrete and let people navigate how they want" in just as condescending a tone and be equally as unconvincing.
There are car lobbyists of course but the streetcars in LA at least were put there by housing developments to sell suburban homes before most people owned cars. Once the homes were sold the corporation that built the rails had no incentive to maintain them, and eventually they were spun off and went bankrupt (of course competing with cars didn't help)
Nor did the fixed price controls they were often saddled with. It seems that politicians are congenitally completely incapable of considering inflation indexing as a concept when they are writing laws.
Street cars are a red herring anyway. Because street cars don't maintain anywhere near the same number of routes as free-form roads. It is a routing problem still, and railed vehicles perform much, much worse at it, which is why they need to be time multiplexed with rail schedules.
Fast forward to blackjack and hookers.
It's tedious to see these same sarcastic comments on every self driving car story. Yes. Buses and trains exist.
When you link the cars together, they usually switch to a hub that's a 10-15 minute walk from your destination instead of your destination and the compartments are occasionally shared with unstable and violent people, which while possibly "efficient" in some metrics, are downsides that many people would rather avoid. Personal compartments are a real differentiating advantage.
violent unstable people aren't inherent to cities.. they're inherent to places that refuse to spend any money on social work/housing/and enjoy punishing people
Slot cars at the grown-up level.
Maybe even put them on steel wheels on steel tracks, to make them more efficient.
Have something like metro (maglev trains) too as they are more efficient than steel wheels on steel tracks.
Public transport for things like metro/trains/trams/buses are honestly underrated.
Rugged American Individualism and Capitalism doesn't allow us to have things like that. We must always be in our individual bubbles away from the filthy poors.
Prior to reading the article, I assumed Waymos were stuck due to an Internet connectivity issue. However, while the root cause is not explicitly stated, it sounds like the Waymos are “confused” by traffic lights being out.
I wonder how Waymos know that the traffic lights are out.
A human can combine a ton of context clues. Like, "Well, we just had a storm, and it was really windy, and the office buildings are all dark, and that Exxon sign is normally lit up but not right now, and everything seems oddly quiet. Evidently, a power outage is the reason I don't see the traffic light lit up. Also other drivers are going through the intersection one by one, as if they think the light is not working."
It's not enough to just analyze the camera data and see neither green nor yellow nor red. Other things can cause that, like a burned out bulb, a sensor hardware problem, a visual obstruction (bird on a utility cable), or one of those louvers that makes the traffic light visible only from certain specific angles.
Since the rules are different depending on whether the light is functioning or not, you really need to know the answer, but it seems hard to be confident. And you probably want to err on the side of the most common situation, which is that the lights are working.
That's what I thought. Then I walked buy Waymos stuck in the middle of the block with nobody in front of them.
I miss the time when "confused" for a computer program was meant in a humorous way.
> miss the time when "confused" for a computer program was meant in a humorous way
Not sure what about this isn’t funny. Nobody died. And the notion that traffic lights going down would not have otherwise caused congestion seems silly.
Not directly. But what about the emergency services not being able to reach their destinations? It stops being funny really fast
> what about the emergency services not being able to reach their destinations?
Did they have documented problems?
This is akin to the Waymos honking at each other at 3AM. Annoying. Potentially dangerous in various circumstances. But ultimately just destructive in a way unlikely to repeat.
Have you seen how human drivers deal with traffic lights and emergency vehicles at the same time? Waymo made the right call to suspend service, they will probably update their playbook to suspend service during power outages in the future.
Humans certainly are imperfect and make mistakes, but will iterate with the understanding that doing nothing at all and blocking emergency vehicles is untenable.
At the least we will fall back to incentive/disincentive social behavior. People will supply ample friendly and unfriendly advice to try to unwind the knot.
Waymo should lose their operating license based on this experience. It's self-evidently dangerous to everyone to be incapable of basic iteration. There's a whole set of law driver's are supposed to follow for handling failed traffic lights. Why have lower expectations of an anonymous car than a human?
> Waymo should lose their operating license based on this experience.
Then everyone should lose their licenses as well by your draconian reasoning. Because…
> There's a whole set of law driver's are supposed to follow for handling failed traffic lights.
And they don’t, it’s chaos.
> Why have lower expectations of an anonymous car than a human?
You obviously have higher expectations for autonomous cars than humans, it is not the other way around for those of us who disagree with you. The only difference is that Waymo can get better with experience and humans generally don’t.
> > There's a whole set of law driver's are supposed to follow for handling failed traffic lights.
> And they don’t, it’s chaos.
Do you live in areas where traffic lights go out regularly?
Because for human driver it is a non-issue. It becomes an all-way stop and you take turns, it is easy. Traffic throughput slows down a bit, but nothing approaching chaos about it. If waymo can't deal with this, that's a problem.
> for human driver it is a non-issue
Genuine question, do we have data for accident rates in traffic-lights-out intersections?
If I remember my research correctly, accident rates go up but fatalities and injuries go way down.
Developing new technologies has risks. In the absence of anything really bad actually happening, I think we can solve the problem by adding new requirements to Waymo's operating license (and all self driving cars) rather than kneecapping the technology.
Same thing as if human drivers have crashed their cars in the middle of an intersection due to traffic lights being out, I would presume.
Would have hoped they trained for this but at least now they likely will be.
frankly at least at the intersection i witnessed i saw plenty of them handling it
That sounds plausible. Humans for the most part can usually navigate that situation to a point. It wouldn't surprise me if Waymo cars weren't even trained for this scenario.
I lived in the training zone for both Waymo and Cruise. They were there for literally years before they were offering rides to anyone. The idea that they could train them for emergency scenarios, especially ones that happen so infrequently like a power outage on a route they regularly drive, seems borderline nonsensical, but I honestly don't know if there is a plausible way to do it.
The one time I saw traffic lights go down, it was total chaos. There were two separate crashes that had already happened when I got there, and there would probably be >1 wreck per few minutes with the driving I observed.
I moved from South Africa to Ireland 2 years ago. It was very noticeable to me how drivers in Ireland have no idea what to do when the lights are out. Absolute chaos!
In south africa, traffic lights not working is a daily occurrence. And we've all learned how to navigate a dead intersection wit zero casualties.
Massive 6 way intersections with 2-4 lanes per direction worked perfectly with everyone taking turns to go.
This is how "science" works in the postmodern world. It's not about predicting, it's about implement, problem, solve.
Waymo aside, it's sad to hear that some parts of the world still have blackouts. Which third-world country does "SF" belong to?
Not texas, that’s for sure.
Interestingly in one of the videos online there are several (five I think) Waymos blocking the right two lanes entering the intersection (along with a few others around the other parts of the intersection). While its hazards are still blinking, one of these vehicles moves forward (admittedly just a few feet).
Is this a violation of the California Vehicle Code? Generally it seems to disallow non-emergency vehicles from traveling with blinking lights except for turn signals (and brake lights responding to a braking action).
There are situations where it is allowed:
https://codes.findlaw.com/ca/vehicle-code/veh-sect-25251/
I couldn't find anything other than their first responders page but IMO any robo taxi operating in a metropolitan area should be publishing their disaster response & recovery plans publicly.
I live in SF in an area that was affected by the blackout. I saw four different Waymos stopped. Three were in the middle of the street. One was along the curb.
My personal opinion. With number of cars I saw flying through blacked out intersection -- major intersections -- I'm very happy that Waymo had a fail safe protocol for such a "white swan"-style event (that is extremely rare, but known-to-happen event).
I saw a damn Muni bus blow through an minor intersection, and was just shaking my head. So many dumbasses behind the wheel, it's miracle no one was killed, and everyone seems to be concerned with "the flow of traffic."
"But think of the thousands that will run late!"
Pretty much no aspect of this event was "extremely rare". PG&E spends the whole year smoking $100 bills and laughing their assess off, then as soon as it rains even a little bit their junk explodes and they pretend they could not have foreseen water existing on Earth. This is not even the first, or second time that this specific substation has burned in living memory. It already burned in 1996 and 2003.
I'm not going to have a semantic argument with you, but I'd consider anything rarer than "a few hours per year" as an "extremely rare event" for the purposes of training autonomous vehicles.
OK well let's argue about the semantics of "autonomous" instead. To me, it means the vehicle's on-board systems should generalize to safe and non-disruptive behaviors under all circumstances. In this instance they should have been able to either navigate to a depot or at least pull off the road.
I think that’s a perfectly reasonable standard for a normal operating environment.
In emergencies I think “safe” is preferable to “non-disruptive.”
I'm surprised that either:
1. Nobody at Waymo thought of this,
2. Somebody did think of it but it wasn't considered important enough to prioritize, or
3. They tried to prep the cars for this and yet they nonetheless failed so badly
Likely 2. Not something that will make it into in their kpis. No one is getting promoted for mitigating black swan events.
Actually that is specifically not true at Google, and I expect it applies to Waymo also.
People get promoted for running DiTR exercises and addressing the issues that are exposed.
Of course the problem is that you can't DiRT all the various Black Swans.
Everyone should have understood that driving requires improvisation in the face of uncommon but inevitable bespoke challenges that this generation of AI is not suited for. Either because it's common sense or because so many people have been shouting it for so long.
What improvisation is required? A traffic light being out is a standard problem with a standard solution. It's just a four-way stop.
In many versions of road rules (I don't know about California), having four vehicles stopped at an intersection without one of the four lanes having priority creates a dining philosophers deadlock, where all four vehicles are giving way to others.
Human drivers can use hand signals to resolve it, but self-driven vehicles may struggle, especially if all four lanes happens to have a self-driven vehicle arrive. Potentially if all vehicles are coordinated by the same company, they can centrally coordinate out-of-band to avoid the deadlock. It becomes even more complex if there are a mix of cars coordinated by different companies.
That only works if everyone else also treats it as a four way stop. Which they don't, unfortunately.
Yes, especially in a city like San Francisco where so many cultures come together: the Prius culture, the BMW culture, the Subaru culture, etc.
I'd say driving only requires not to handle uncommon situation dangerously. And stopping when you can't handle something fits my criteria.
Also I'm not sure it's entirely AI's fault. What do you do when you realistically have to break some rules? Like here, I assume you'd have to cut someone off if you don't want to wait forever. Who's gonna build a car that breaks rules sometimes, and what regulator will approve it?
If you are driving a car on a public street and your solution to getting confused is stopping your car in the middle of the road wherever this confusion happens to arise, and sitting there for however long you are confused, you should not be driving a car in the first place. That includes AI cars.
In practice, no one treats it as a four-way stop, which makes it dangerous to treat it as one.
Drove through SF this evening. Most people treated it as a four-way stop! I was generally impressed.
To be fair 'common sense' and 'many people have been shouting it' about technical matters have a long history of being hilariously wrong. Like claims that trains would cause organ damage to their riders from going at the blistering speed of either 35 or 50 mph, IIRC. Or about manned flight being impossible. Common sense would tell you that launching a bunch of broadcasting precise clocks into orbit wouldn't be usable to determine the distance, and yet here we are with GPS.
But a citywide blackout isn’t that uncommon.
> But a citywide blackout isn’t that uncommon.
I think too many people talk past each other when they use the word common, especially when talking about car trips.
A blackout (doesn't have to be citywide) may not be periodic but it's certainly frequent with a frequency above 1 per year.
Many people say "common" meaning "frequent", and many people say "common" meaning "periodic".
Even among people who mean "common" as in "frequent", they aren't necessarily talking about the same frequency. That's why online communication is tricky!
It isn't? To me that's the main problem here, as this should be an exceptionally rare occurrence.
I think that statement is regional. I’ve never seen one.
[dead]
Clearly cars can navigate themselves, it's the lack of remote ops that halted everything
Waymo's performance in this outage was horrible. 6 hours into the blackout there were still many intersections where a Waymo was blocking traffic, unable to navigate out of the way. This should never happen again.
To get permit to operate in cities, Do these companies submit the list of edge-cases they handle?
Each city will have its own nuances.
Why don't the regulators publish the list?
This was very annoying, and made things feel unsafe. Having vehicles stopped blocking visibility when there is no light. Its bad enough we tolerate them stopping and waiting for a pickup and blocking lanes under normal conditions. I had a hard time seeing if there are pedestrians when they’re literally in the cross walk stopped.
"the robotaxis are reliant on infrastructure out of the company’s control"
Well there's your problem.
I don't understand why everyone is talking about the cars when the bigger issue is why the critical infrastructure (lights) don't have batteries for backup.
Obvious failure should be obvious. Get out of the streets. What happens when one gets a flat tire? Surely it doesn't just stop in the middle of the street, right??
Anybody on the ground confirm if it was the traffic lights or lack of cellular that cussed the stoppages?
I saw plenty of Waymos managing to make it through intersections. They were slow and tentative, but definitely made forward progress.
I think the emergency "phone home" protocol requires a phone, presumably with enough channel capacity for reasonable video feeds. I wouldn't be surprised if the dead in the road Waymos were lacking connectivity.
There is of course also a possibility that the total demand exceeded the number of people at Waymos available for human intervention.
I think it’s clear that both use cases are a must have during an emergency. Even more, rescue services and stranded people would need all the bandwidth and reception they can get, Waymos shouldn’t be online during such times at all.
First responders get the highest priority of cellular networks even in non-emergency situations (FirstNet).
FirstNet is AT&T. Verizon and T-Mobile have their own alternatives (FrontLine and T-Priority, respectively). QCI is a fun rabbit hole, because some of these networks share the first responder QCIs with certain business use cases.
I didn’t notice a lack of cellular. Though it did get down to like 6Mbps, which was certainly degraded service.
I was in the affected area and we effectively lost all but messaging. Not the whole time, but definitely while I was ordering takeout at a place with power. I couldn't get an image to send to a friend.
I lost cell during the whole outage on Verizon, came back immediately when power was restored. There seemed to be some towers up, if i walked down the street I could find one, but plenty were down.
How did FSD Teslas do at the traffic signals? Or Nuro?
Here's an example of it doing well:
https://x.com/MarioNawfal/status/2002699317409186219
The Tesla example is a single oncoming car, clear right of way, and ample time in a simple 4-way intersection.
The Waymo video has over a dozen cars, at least 6 pedestrians crossing streets (many more on the sidewalks), and is a 5-way intersection.
These are cherry picked examples. Either advertising or propaganda.
That's a really good question.
Waymo should do a bit more research in reliability and explainability of their AI models.
Honestly, I’m just glad this stopped before a major accident happened.
More discussion:
PG&E outages in S.F. leave 130k without electricity
https://news.ycombinator.com/item?id=46342022
Did anybody get stranded on the bart?
No, but they ran through stations that lacked power, so in the sense of being dumped out far from your actual destination, yes.
Not looking forward to having this junk clogging up my city.
It was predicted by many, including me. It'll be a lot worse in an earthquake where power and cell service are out and there's debris and road damage. Good luck to our first responders.
https://news.ycombinator.com/item?id=41688847
How does Tesla FSD respond to inactive traffic control lights?
Coincidentally we were on the Robotaxi during the black out (didn’t know about it, we were going to Japan town from the Mission). Noticed that it navigated through the non-working traffic lights fine, treated it like a stop sign junction. One advantage of building unsupervised system from public version that had to deal with these edge cases all around the country.
Though the safety driver disengaged twice to let emergency vehicles pass safely.
Treats it as a four way stop.
https://www.tesla.com/ownersmanual/modely/en_eu/GUID-A701F7D...
[flagged]
https://x.com/edgecase411/status/2002630953844552094
Looks like it treats it as a 4-way stop. Is this because Tesla has more training data?
I'd default to assuming it's the respective roadmaps for Waymo and Tesla differed on which things to implement when, not training data, that results in the two behaving different.
50/50 bet it would either go right through or treat it as a stop.
Don’t think I have had a totally inactive light. I have had the power is out but emergency battery turned to blinking red light, and it correctly treats as a stop sign.
> Is this because Tesla has more training data?
Its human takes over. FSD is still Level 3.
(Robotaxi, Tesla's Level 4 product, is still in beta. Based on reports, its humans had to intervene.)
FSD is level 2. Level 3 doesn't require the human driver to monitor the outside environment, only take over when requested. Tesla also doesn't report data from FSD under L3 reporting requirements anywhere in the US.
[flagged]
The proof you all needed that these Waymos were teleoperated all along.
I thought this was trivially verifiable due to regulation and latency analysis.
If they truly rely on teleoperation, that's at least 20ms in the best case, and can grow a lot with interference.
I always assumed these things have some autonomy.
I for one welcome our robot slow-verlords.
I thought LIDAR solves for this lol
[flagged]
If you think this is swell, just wait until they move us to 100% digital currency!
As a society, as a whole society, opposed to our narrow interests and point of views as typical affluent HN dwellers, do we really NEED this kind of shit?
I think I prefer trains and buses.
SF Muni & BART both stopped service in many areas. Though most of the trains still had electricity, many sensors and control systems were inoperable. Also underground stations had no lighting, so it would be hazardous to allow people to board or exit there.
Waymo's problem is obvious in hindsight, and quite embarrassing for them, but it can be solved with software improvements. Tesla's FSD already treats dark traffic lights as stop signs, so I would bet on Waymo fixing this as soon as they can.
But transportation that depends on infrastructure along the whole route (such as trains and busses powered by overhead lines) are always going to fail in these situations. I think that's acceptable considering how rare these events are.
Living in social rot and keeping unmanned little autos for those that can afford it seems even more nasty than what I initially had in mind.
It seems waymo's always fall apart when encountering something that wouldn't be in the training set. Such as a christmas parade:
https://www.youtube.com/shorts/NOqK8UEuWjs
I don't understand how these cars keep getting stalled for half an hour or something. Surely there must be a team of teleoperators ready to jump in at any time?
The power outage probably knocked out the infra those operators needed to control the cars.
True but the comment I replied to mentions a different case where a Waymo got stuck for half an hour at a parade.
"the infra" is cell phone data coverage though. Which was probably congested by all the city residents using theirs instead of their wifi which was down. Would be fascinating to see just how much Internet traffic flows changed during the outage.
In my neighborhood, Xfinity goes out every time we lose power.
Even with my generator and UPS powering my modem, the outage is not resolved until mains power comes back
These monopolies aren’t required to have uptime in the same way the POTS network was.
We had a week long outage last year and people were driving 15 minutes out to get cell signal to catch up on their data
Until AVs can deal with OOD scenarios, they should not be on the road.
Autonomous vehicles should be on the road iff they reduce overall incidents/deaths. Failure to deal with an out-of-distribution scenario would count against this, but may be rare enough to not significantly affect the average.
So basically no answer from Waymo other than to boast about their numbers? Why not just be transparent?
The "boast about numbers" wasn't from Waymo, it was from their investors, and it was earlier this month and not during the service suspension.
What do you mean no answer? They are suspending their operations. That is their answer.
Read more carefully, Waymo didn’t boast about their numbers and that part of the story was unrelated to the issue except as a measure of impact their shutdown could have.
Because they are currently trying to raise money at 100bn valuation
"San Francisco Mayor Daniel Lurie warned residents to stay off the roads unless they needed to travel."
Sure, I go out and drive around on the roads for no reason all the time. I'll avoid doing that during the crisis.
So maybe the point was to consider a different store or route.
Or look at the traffic and decide if you REALLY want to spend an hour or more in gridlock for whatever activity you are considering.
And maybe wherever you wanted to go is closed because they don't have power either.
It is a perfectly reasonable request.
The fact that you acknowledge that is was a "crisis" implies pretty strongly that you understand that a priority evaluation might be useful.
The word carrying water here is "need." My partner and I ran an errand in the blackout, we didn't need to do it, but we wanted to. I drove us almost exclusively through an area of the city that had power, but we still ran into issues, including multiple disabled Waymos during that trip. Even in the areas with power, people were driving weirdly. If were able to do it over, I would have stayed home.
I did when I was younger (18-25). Exploring the world outside of my hometown often put me at ease when I was feeling destabilized.
I used to go out in the snow on purpose (DC metro). Mostly have them to myself, it was always fun.
Unsure if sarcasm, but lots of people do that. I do it every day.
Some folks enjoy driving.
Yeah it was sarcasm. I enjoy driving also, actually. But I don't just go out and drive for no reason. It's always with a purposeful destination.