Creating an all-weather driver

https://waymo.com/blog/2025/10/creating-an-all-weather-driver

by boulos • 9 months ago

125 106 9 months ago

daft_pink• 9 months ago

i’m really curious at what point it decides that it shouldn’t be driving.

Humans are horrible at this I wonder what the limit is. I've always thought that I can tailor my speed to conditions but not everyone on the road slows down.

hangonhn• 9 months ago

It's really interesting because that's something they definitely don't teach you when you first learn to drive. Growing up in Florida, I learned to pull over and turn on emergency blinkers if the rain gets bad enough. The reason I know to do this is because I saw other drivers do this on the highway and realized that's pretty wise. It's tempting to imagine that a younger version of me would have been smart enough to realize this on my own but I think most of us learn a lot by observing the behavior of others. Or maybe I would have learned eventually after a few close calls with skidding. Or maybe I would have never learned until it's too late. I wonder if the different responses to averse conditions you've observed is a function of the different experiences we've had as drivers. You might be a more experienced driver than some of those around you.

amluto• 9 months ago

Humans have one advantage over autonomous cars in ice: they can pull over and put on chains. Cars can’t do that (yet).

(I’d love to see a serious winter vehicle that can deploy traction devices by itself, perhaps while rolling at very low speed. Off the top of my head, it seems like it might be easier to put them on then to take them off.)

rangestransform• 9 months ago

The limit is much higher than human performance given enough low latency compute. [1] is probably the limit, the actual issue is being able to do that while also avoiding colliding with other road users. The challenges of state estimation and control should be the same.

[1]https://www.youtube.com/watch?v=2MWib6bnnT8

potato37328428• 9 months ago

I hate this sort of take. Humans aren't horrible, on average, they're about average. Your opinion is just a statement of "my judgement is decently far from the fat part of the bell curve" but dressed up in "and I know better" type snobbery.

url00• 9 months ago

Exactly. I picture a dystopia where the car refuses to attempt escape from a storm because of the liability factor.

brookst• 9 months ago

Sounds preferable to a dystopia where AI driven cars are getting into wrecks because they’re overconfident in their abilities.

dingnuts2• 9 months ago

I picture one where it locks the doors and drives you right to the ICE center as soon as the facial recognition cameras realize who you are

even better if this is the only way to get around. no transport for whoever the Trump admin decides is insufficiently loyal!

y'all need to get more creative with your dystopias

precommunicator• 9 months ago

I see no issue with that as long as you can override and drive manually.

dingnuts2• 9 months ago

when the remote operator watching five feeds notices it's doing something dangerous

milleramp• 9 months ago

At the Los Angeles Ciclavia two weeks ago Waymo's were getting stuck at the car crossings. There were police standing there waving cars through but the two I saw were not willing to drive through the intersection.

nradov3• 9 months ago

Properly responding to informal hand and voice signals from law enforcement, road workers, and other humans is going to be one of the toughest technical challenges for autonomous vehicles to solve.

darth_avocado5• 9 months ago

Stop signs became universal. No reason why machine readable signals/devices to communicate don’t become the norm with law enforcement and emergency response workers.

dghlsakjg• 9 months ago

Quite frankly, many drivers don't do well here either since hand signs can be very ambiguous. And many times there are contradictory signals that require interpretation.

Look at Scottie Scheffler's arrest for an extreme case of how very hard this is to get right.

tonymet4• 9 months ago

I hope this improves rigor and common sense around winter driving in the USA. In Eastern Europe, drivers care more about tires, angility and driver skill. In the USA , drivers rely on large 4wd vehicles with high clearance for snow and ice driving. I’ve seen way too many issues with large clumsy vehicles losing control due to poor tires .

I hope Waymo shares more solutions for winter driving to debunk a lot of the marketing for winter activity driving in the USA

micromacrofoot2• 9 months ago

It won't, our economy is somewhat reliant on giant vehicles that people can barely afford to maintain.

randerson• 9 months ago

Manufacturers should fit all-weather tires by default (not all-seasons) - they are decent in both summer and snow (3PMSF).

The average car owner seems oblivious to the different types of tires. Most high performance cars come with summer tires. I live in a wealthy area where I often see new cars in parking lots wearing summer tires in winter, probably relying on electronic nannies to mask the lack of grip in normal driving.

ghaff6• 9 months ago

Many large 4wd vehicles are nothing special with respect to ground clearance which mostly doesn't make much difference for snow/ice driving on paved roads anyway.

chemotaxis2• 9 months ago

I don't think the cultural difference you're describing here really exists. Maybe if you mean people from the SF Bay Area who visit Tahoe. If you go to places with real winters, people know about winter / studded tires, will often carry chains, and so on.

tonymet4• 9 months ago

some truth yes. even where I live with plenty of winter conditions, less than the midwest, still lots of poor car and tire choices. 6K# SUvs. even in the midwest lots of huge vehicles. perhaps with better tires, but still impractical.

danielmarkbruce• 9 months ago

100% this. It's laughable how many times europeans make sweeping generalizations about the US. There are various places in the US where it snows rarely and yeah, people (including me) are clueless when it happens. And then there are people in Buffalo who are more than capable of handling the snow.

chermi• 9 months ago

(citation needed)

b0rbb• 9 months ago

Upstate New York

I'm guessing they meant _Upstate AND Western New York_.

Glad someone in Waymo saw the potential for testing for extreme snowy conditions there.

boulos• 9 months ago

Yes. We went to Buffalo, and a few other locations (https://techcrunch.com/2024/08/15/waymo-to-double-down-on-wi... and other reports)

umanwizard• 9 months ago

Anecdotally I feel like the Upstate vs. Western NY distinction is mostly only made by people who live there.

When I lived in NYC I used "upstate" to mean anything not in the five boroughs, Long Island or Westchester, and I don't think this usage is uncommon.

counters• 9 months ago

Eh, it's a pretty big distinction weather-wise. Extreme Western New York and the Tug Hill plateau are all susceptible to somewhat frequent lake effect snow. Given the right time of year and wind fetch, you can see narrow convective / lake-effect snow bands from the Finger Lakes. But broadly speaking the actual annual expected snow and the phenomenology of the storm systems that produce that snow are very different over the rest of the state.

awaymazdacx5• 9 months ago

dragnet which is LPRs for fleet vehicles

marstall• 9 months ago

boston: the ultimate test

ghaff6• 9 months ago

Had to drive someone to the Fenway area the other day. And that was bad enough in perfectly reasonable weather :-) I'm OK with driving into the cit(ies) in general but don't regularly go into that area of town.

potato37328428• 9 months ago

It's easier in bad weather because the ~10% of stupid people who traffic normally just kinda flows around slow it to a traffic jam that's pretty impossible to screw up too badly.

darth_avocado5• 9 months ago

I was told by a very intelligent man demanding a trillion dollar salary that you only need vision cameras to have full self driving in all weather conditions. All of this is apparently unnecessary.

TrainedMonkey• 9 months ago

He is not wrong, but we demand superhuman performance from our machines which in this case necessitates superhuman sensory abilities. Current evidence shows that having non-vision sensors is a faster way to create a reliable system. I would personally choose to ride in an autonomous vehicle with Lidars.

Gigachad• 9 months ago

It seems quite likely that once self driving cars are well perfected, we will demand more than just human level driving which is currently horrendously dangerous. If lidar systems can exceed vision only, we are going to demand it as a baseline standard.

dbt002• 9 months ago

He is not wrong, but we demand superhuman performance from our machines

I have a model 3 with v3 FSD hardware. FSD is an objectively terrible driver compared to the average human.

UltraSane2• 9 months ago

He is very obviously wrong since Waymo cars drive millions of trips with 0 drivers while every single robotaxi still has a safety driver in it at all times.

CamperBob2• 9 months ago

"He is not wrong, but anyway, here's why he's wrong."

He's wrong. Cameras are not enough... but they're certainly cheap enough.

tanseydavid• 9 months ago

The vision-only approach surely seems to be falling behind the multi-sensor approach.

tom13372• 9 months ago

The vision-only approach isn't even able to trigger the windshield wipers correctly.

loeg5• 9 months ago

It's fine to dislike the guy, but stock options aren't the same thing as a salary.

Pwntastic• 9 months ago

they are when you can open a line of credit against them

darth_avocado5• 9 months ago

Sure. But when the option is priced at strike price way below the market price, it can be pretty much considered as salary.

jillesvangurp• 9 months ago

Lots of people manage to drive in poor visibility without radar, lidar, etc. If that's safe and normal for people, the principle that self driving should work with just cameras isn't such a strange one.

I wouldn't call things like radar and lidar unnecessary but in principle a good AI vision system should be able to operate at the same level as a human eventually. Of course if you don't have such an AI just yet, you need a stop gap solution. But I wouldn't bet against AI getting there eventually. Probably not on Mr. Musk's accelerated and optimistic schedule though. But give it another five to ten years and things might look a bit differently.

What Waymo is doing now with much less than perfect AI is of course completely pragmatic and very impressive. I'm kind of eager to see them start operating self driving outside a few restricted zones in the US and for example in European cities. I live in Berlin, so probably we'll have to wait quite a bit for people to finally let go of their fax machines though apparently there are some trials with self driving buses about to kick off here now.

Zanfa2• 9 months ago

Lots of people manage to drive in poor visibility without radar, lidar, etc. If that's safe and normal for people, the principle that self driving should work with just cameras isn't such a strange one.

No camera system comes close to the capabilities of human eyes, combined with general intelligence.

bluGill• 9 months ago

I am one of those who drive in bad weather from time to time. I'm 'good at it' - but I cannot honestly call myself safe. I've been in the ditch. I've spun a 360 and only didn't hit someone else in the process because it happened nobody was there.

i grew up where bad weather was common enough that we cosidered it not worth shutting down for bad weather so we risked driving in it - but it was always a risk and many do die from taking that risk.

reorder9695• 9 months ago

Just vision is a very reductive way of describing how we drive. We use sound to hear if something else is coming and where from, or if we're losing traction on the road, or if the road surface has changed. We can judge acceleration and deceleration in all directions using our inner ears. We can feel if the car's performing differently to usual indicating an issue with the car or different conditions. In addition when we are using our vision, if it gets obscured (i.e. snow covers the windscreen) we know how to get it off with wipers, and most importantly we're very adaptable to new conditions in a way that computers aren't, if we experience something completely new to us, chances are we'll make a reasonable in the moment decision.

Theodores• 9 months ago

I don't think either solution is going to be the eventual winner. I expect the eventual rollout of 5G+ to be the game changer.

Much like how a pilot captain boards a ship to steer it into port, traffic systems will be able to be a bit more 'hands on' when it comes to getting traffic through junctions safely, regardless of the weather. Hence, on the final journey through a city, the city traffic systems will be directing your car.

For the journey on highways between built up areas, variants of today's self driving systems will suffice.

zeroq3• 9 months ago

I remember, back in the day, when first driving assistance systems rolled out - you know, keep lane, speed assistance according to road signs, etc. - I thought to myself "I bet you haven't seen our roads".

When I was getting my driver license I had to perform a series of tasks as part of the process. On of first was driving a 50m narrow curve forth and back. I had my exam in the middle of winter. The training yard was fully covered in snow. I was young and didn't knew better, so I got in the car and begin my test to quickly realize I couldn't see shit. I tried my best, but next moment I was told I got off the curve - the examiner knew it by heart - and I failed the test. Back to school and see you next semester.

A learning experience on so many levels.

Anyway, since then I always come back to that single experience when I read about self driving vehicles.

voidfunc• 9 months ago

Heh this must be Europe. In the US the driving test I took involved taking a right hand turn. Another right. Execute a 3-point and then drive back the half a mile to the testing center.

zeroq3• 9 months ago

Correct.

You have to pass theory exam - when I was doing it it was 30 questions with four possible answers out of ~500 pool [1] now the pool is 3700. If I'm not mistaken you could make two mistakes and still pass the test. The questions are either road diagrams, ie. intersection with three cars, road signs and/or lights and you have to tell in which order cars will go or a picture from drivers POV and a question what you should do in such situation - like three lanes, car on the middle, the right is a bus lane and the question is if the driver is allowed to take the left lane.

Once you get that done you can take practical driving test.

It starts on a training yard where you have a series of tasks, like said driving forth and back on a curve in one sweep motion, starting a car on a incline without going back or losing your engine - mind you we're driving manual - parallel parking between cones [2], etc.

If you ace them instructor will take you for a 30-45 minutes ride around town. Apart from normal driving he will ask you to few random tasks like parking in normal conditions.

Any mistake will end the trial with a fail and you have to start all over again. Three failures and you need to redo your theory exam.

[1] I have really good memory and when I taking my exam I went through all questions few times - they are publicly available - and when I had my exam I only looked at the picture and double checked if the question ended up with question mark or period as some images were reused and I knew my answer. I don't think I would be able to do that again with that 3700 question database lol.

[2] before taking exams you go to a private driving school and my instructor gave me a cheat code for parallel parking - which is extremely tight, but also, as pointed by my instructor, government regulated, so all the cones have to be in very precise spots. Not only that, but you take the exam in government selected car (whoever won current bids). So he told me to back up until I saw a cone lining with door post, then full stop, rotate the steering wheel by a exact amount of degrees, etc.

PS. gun permits are given on similar grounds, plus you are required to have a regulated gun safe at your premise and it has to be permanently attached, so given the fact that many of our apartments are smaller than your garages and most people are renting significantly reduces access to guns :)

dmd• 9 months ago

I (1997, upstate NY) didn't even do that much. I got in the car with the instructor, we drove around the block once. I ran the stop sign because a truck was parked in front of it. The instructor didn't notice (I only noticed later).

potato37328428• 9 months ago

I know, I know, murka bad, but my driving test decades ago had a left turn out of a driveway, a 3-pt on a road with real traffic, parallel parking, and an intersection with right on red to return to the testing center, located in a terrible parking lot (i.e. designed in accordance with modern standards) that had a ton of stupid lines that "normal people" cut but you have to follow as part of the test.

gniv• 9 months ago

I wish Google would use their learnings from Waymo/Streetview in Maps navigation. When I drive through a complicated intersection for the first time it's a bit of guesswork what's the right thing to do. Here in France at least since space is at a premium there are a lot of weird intersections that are hard to navigate.

okdood64• 9 months ago

They're a separate company.

tim333• 9 months ago

I've found maps navigation can be quite good with complex junctions in places like Austin where there can be seven lanes splitting off in weird ways. Maybe they are just confused by French junctions as I have been myself at times. The Arc de Triomphe is quite a thing to drive around. (as in https://youtu.be/-2RCPpdmSVg)

IneffablePigeon• 9 months ago

What learnings? Google maps can’t even reliably tell me what lane to be in for the next junction, or what the speed limit of the current road is.

Creating an all-weather driver

Related Stories