Waymo updates 3,800 robotaxis after they 'drive into standing water' • Hacker News Board

Re: Waymo updates 3,800 robotaxis after they 'drive into standing water'

#151

Post by pfdietz » Sat, May 16, 2026, 12:51 AM UTC

tialaramex wrote: post #121

drob518 wrote: post #22

Earlier quoted context omitted.

It’s a particularly hard problem in Texas. We get torrential rains and the landscape is relatively flat. Couple that with shallow soil over lots of limestone and it means flooding is really common. We also have roads that have a “low water crossing,” where a road crosses a creekbed that is normally dry but which will flood. There are often water depth signs there (basically a vertical ruler with feet marks so you can…

Yeah, people are bad at guessing and the usual "Plan continuation bias" kicks in. I was travelling in a group to eat lunch with friends once, after heavy rains. We reached a site where the road needs to fit under a bridge and is known to flood, there's standing water, and the driver figured it's probably not too high, he drives in and nope, water over the air intake, bye bye engine and we walked the rest of the way t…

Uncle Rodney has left the crankcase.

Re: Waymo updates 3,800 robotaxis after they 'drive into standing water'

#152

Post by hansvm » Sat, May 16, 2026, 1:12 AM UTC

If we're lucky, this'll help get us better roads. I'm never exactly happy slowing down to 10mph on the freeway because it isn't obvious whether the 100ft long puddle is an inch deep or 2ft, and I can sidestep some of that danger sometimes by proactively choosing safer lanes, but our roads are dangerous and don't handle water correctly.

Re: Waymo updates 3,800 robotaxis after they 'drive into standing water'

#153

Post by ChoGGi » Sat, May 16, 2026, 1:16 AM UTC

Chaseraph wrote: post #134

drob518 wrote: post #22

Earlier quoted context omitted.

It’s a particularly hard problem in Texas. We get torrential rains and the landscape is relatively flat. Couple that with shallow soil over lots of limestone and it means flooding is really common. We also have roads that have a “low water crossing,” where a road crosses a creekbed that is normally dry but which will flood. There are often water depth signs there (basically a vertical ruler with feet marks so you can…

This case makes me think of my brother's place in rural Tennessee. To get to his house, you drive through a small creek, year round. For a hundred years in their community, they've managed without a bridge. I'm not sure driverless cars are ready for edge cases like this. Also, no one tell Enterprise I drove their rental through a creek.

Heck, that was the way I took into the city for work for a few years, shaved a good 30 mins off the commute.

You'd have to hold off for a few weeks every season change while the ice hardened up/melted or get stuck in it (thankfully I tended to get there after someone else found out).

Re: Waymo updates 3,800 robotaxis after they 'drive into standing water'

#154

Post by ChoGGi » Sat, May 16, 2026, 1:20 AM UTC

dietr1ch wrote: post #29

tintor wrote: post #16

Earlier quoted context omitted.

Such detailed database of fine grained road geometry gets stale very quickly, due to road maintenance and road construction. In US highway lanes are shifted sideways frequently.

But are they not continuously updating the road database with their fleet?

Earthquake or sinkhole?

Re: Waymo updates 3,800 robotaxis after they 'drive into standing water'

#155

Post by ajsnigrutin » Sat, May 16, 2026, 1:23 AM UTC

marvinkennis wrote: post #144

Animats wrote: post #2

That's a tough problem - distinguishing wet pavement from deep water. Humans make that mistake frequently. Autonomous vehicles should probably be equipped with a water sensor. (We did that in our DARPA Grand Challenge vehicle back in 2005). Then they can enter water very cautiously and see if it's too deep. This may make them too cautious about shallow puddles on roads, though.

They should really just park themselves on the side of the road (or observe in real time), wait for another car to go first, then follow that path

It would have to know the height of that car above ground, how high the air intake is, etc. A lifter offroader could make it through a much deeper body of water than eg. a prius.

Re: Waymo updates 3,800 robotaxis after they 'drive into standing water'

#156

Post by cma » Sat, May 16, 2026, 1:27 AM UTC

nomel wrote: post #131

harry8 wrote: post #83

Earlier quoted context omitted.

So the first waymo to get to this less used road to john’s will not have the data rather than every waymo that travels down a new highway, that then becomes a problem if it rains. One car with an issue of first coincides with rain on a less used road?

Well, it's closer to: any car with stale data and sufficient water depth is a financial and PR disaster. These cars are not cheap, and a tiktok of someone being driven into the water is even more expensive!

As soon as the car descends below what was mapped it should be able to know there is a discrepency.

Satellite monitoring is also available for detecting extensive road work which they could use to invalidate and send out something to remap.

Re: Waymo updates 3,800 robotaxis after they 'drive into standing water'

#157

Post by cma » Sat, May 16, 2026, 1:30 AM UTC

kpw94 wrote: post #21

wombat-man wrote: post #7

Earlier quoted context omitted.

If they have a laser measurement of the road from before, couldn't they see that the level of water vs the expected road surface?

That seems a very risky assumption for any car (self driving or human driver) during flash floods. "Turn around don't drown": You think you know how deep it is under because you've taken that road many times before (or in your case you have historical laser measurement) But you don't know: - Maybe the road under fully collapsed - Maybe the flow of water is extremely strong, so you need to accurately estimate that too…

Flow should be able to be done with vision, radar can as well: some bridges use surface flow monitoring radar.

Re: Waymo updates 3,800 robotaxis after they 'drive into standing water'

#158

Post by cma » Sat, May 16, 2026, 1:31 AM UTC

flutas wrote: post #55

filoleg wrote: post #46

Earlier quoted context omitted.

Granted, I am not sure exactly how Waymo operates, but I thought that the extensive testing was mostly for legal reasons+just handling edge cases. I am saying this, because I noticed that they typically start with a low-tier restrictive permit to operate (with a rather small number of cars in the fleet). Then they run it for a year or two, iron out edge cases particular to a given city (e.g., climate particularities,…

Waymo explicitly lidar scans and "HD maps" the area: https://waymo.com/blog/2020/09/the-waymo-driver-handbook-map... Tesla is less "HD", they have standard maps like we all think of, and a lane level "see-ahead" system where they basically just grab a satellite image tile, and align it with what the car sees for "FSD".

Some of the Tesla robotaxi deployments have relied on HD maps too, or at least they were spotted extensively lidar mapping Austin.

Re: Waymo updates 3,800 robotaxis after they 'drive into standing water'

#159

Post by 1970-01-01 » Sat, May 16, 2026, 1:34 AM UTC

Animats wrote: post #2

That's a tough problem - distinguishing wet pavement from deep water. Humans make that mistake frequently. Autonomous vehicles should probably be equipped with a water sensor. (We did that in our DARPA Grand Challenge vehicle back in 2005). Then they can enter water very cautiously and see if it's too deep. This may make them too cautious about shallow puddles on roads, though.

I thought the same thing. A very small float switch would work here. Somewhere between the radiator and the bumper. Fording depth is different for every vehicle.

Re: Waymo updates 3,800 robotaxis after they 'drive into standing water'

#160

Post by doright » Sat, May 16, 2026, 1:48 AM UTC

Animats wrote: post #2

That's a tough problem - distinguishing wet pavement from deep water. Humans make that mistake frequently. Autonomous vehicles should probably be equipped with a water sensor. (We did that in our DARPA Grand Challenge vehicle back in 2005). Then they can enter water very cautiously and see if it's too deep. This may make them too cautious about shallow puddles on roads, though.

Reminds me of all the Waymo vehicles stalled during that San Francisco blackout a while ago.

I have always believed that when people cite statistics on Waymos beating human drivers on safety statistics, that is only in the case of the happy path, or "happy road". The safety statistics could plummet in specific scenarios that lack training data or forethought, and they could crop up at any time.