Hacker News Reader

Comment by wombat-man | Hacker News Reader

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?

View on HN

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?

For common routes, yes. For getting to John's house, where the path there sometimes floods, no.

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.

Sure, if you drive around slow enough so you can stop in time. Lets say coefficient of friction is around 0.5. That means you can drive around town at a brisk 12mph, if you need to stop within 10 feet (with 0ms reaction time).

People usually drive slow too through flooded sections. They don't have the advantage of any descent of the car diverging from HD maps, or any water level visually below HD map surface serving as an immediate warning.

> drive slow too through flooded section

This first requires knowing you're in a flooded section. That's the gap here, is it not? My point is that you can't use descent AS a "flooded zone" detection because then your speed, by the harsh mistress of psychics, is very limited everywhere you go, flooded or not, because stopping distance has to be kept very short since just a few feet of overshoot, when your flood detection triggers, is the difference between ruined engine or not.

Maybe that's their fix, if a flat mirrored surface is detected, slow way down, because perceiving depth of muddy water before getting in it is hard.

I'm assuming there is no issue detecting water over the road, only depth below it that has changed since the HD map was made.

And, we're full circle at this point.

They had an issue detecting them here. But my understanding was people were saying beyond seeing the flooded section there is a fundamental problem that they can never solve that if there is new construction under the flooded section lowering the level of the road they can never deal with that. But as long as it is not a sudden dip or sinkhole I think HD mapping can still help with that since they can monitor the cars altitude as it traverses it and detect divergence from what is mapped early.

Earthquake or sinkhole?

I traveled to Austin 3 weeks ago and there were entire highways not on Google Maps.

Apparently they were built in just a few months.

There's some places where Apple still thinks I'm driving through a cornfield even though the development is a few years old, now.

I suppose I could inform them somehow, but it's not worth the bother.

I don't know how they don't notice thousands of users driving through these "cornfields" at 60mph every day, though. You'd have thought that'd raise some alert?

I’ve literally watched my car (only car around) trigger the yellow heavy traffic warning (because I was driving slow to look for something).

Yet thousands of cars doing 60 mph through a cornfield and over a river doesn’t trigger a “maybe a freeway was built here?”

Advertising slightly heavier traffic isn't the worst thing in the world. Telling hundreds of motorists "There's a navigable road here" when there isn't, is a bigger problem, for both Apple's PR team, and the motorists in question

Verifying indepenantly is probably the reasonable response

Seems like that would be easy to do with some current satellite imagery?

Waze was pretty public about how they used it, with automatic flags and route updates with no/little human intervention. I assume everyone else is doing the same, since it's so obvious. I'm sure they know, but if it was a cornfield a couple years before, there's a good chance it's lower on the priority list, in a "how many people will this impact" sense.

I’m still amazed at the people who claim that Apple Maps is as good as Google Maps nowadays. If you live anywhere where there’s lots of development, it’s definitely not. It’s also terrible when businesses or places of interest move. My wife’s business moved a half mile down the road and a single message to google maps got it moved in a couple days. Apple Maps took about a year with multiple requests and even multiple messages to their special “escalation” email address.

Agree that Google does a much better job of pulling in latest road and business info, even in the US, even in California, even in Cupertino.

But if you’re not going to some brand new addeess on a road that didn’t exist six months ago, Apple maps are just so much more readable and usable. Sucks thst the fata isn’t as good, but damn does it show what a difference colors, fonts, and tasteful selection of what to show and what to hide nakes,

I routinely have to drop into Google maps (international travel, and some newly developed areas) and it always feels like time travel to a 90’s website. Except for the whole “data is current” thing.

Pretty sure they already rely on such a database for positioning, so they already have that problem.

But yes, this wouldn't work for other self-driving systems that don't rely on HD maps.

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.

I more meant that it could maybe see a significant difference in the road, and know to take caution, not to try to gauge the depth of a submerged roadway.

You underestimate how frequently details like this change in the real world and how difficult it is to reliably integrate them into the mapping models with very low error rates.

Aggregating this data in something close to real-time, verifying and corroborating that the change to the road model is real and correct, and then pushing those model updates to every vehicle that may need it almost immediately is not really a solved problem.

That's so much extra complexity

If they have a pre-existing database of every road, sure. And if it's kept up-to-date at all times in all vehicles.

Isn’t that the Waymo data model, though? They extensively pre-drive every new market, building dense volumetric maps of the entire service area before they begin service, so they essentially do have that database of every road (that they drive on).

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, crazy spaghetti junctions in ATL, etc.), and log a lot of data. Then they take that data, go to the city/state, say "we have all this data that demonstrates we were very above the board while running the test pilot program, we are safe, and now we want to expand out of a very limited test pilot program."

And then it either goes well (Bay Area, LA, etc.) or goes off the rails for other reasons (often failing earlier for entirely unexpected reasons, like the pushback against it from taxi driver unions in NYC).

My point being, I could be entirely wrong, but I don't think that they literally map every single inch of the road before being allowed to operate. I just don't see it as being possible in any large populated city, given how often things change there. Just in 3 years living at one apartment in Seattle, I had a road directly adjacent to me changed from 2-way to 1-way, and then had 3-4 lanes that were basically highway entrances/exits (a block away from me) created and the whole area being rerouted entirely.

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.

Those vehicles with lidar roof racks are validation for "FSD" look up project rodeo for some news reporting (they also have cameras further up mounted on a pole to validate "FSD"s reprojection technique).

Images of the cars screens show "project rodeo controls" is the only reason I say that we know these are validation vs mapping.

Waymo does high resolution scans, the question only they know the answer to is how well does their model do without them and on camera only. I bet it's way better than is publicly acknowledged.

They actually do significant mapping. Where it operates currently it is not unusual to see this. It will be a waymo with a human driver operating someplace not currently in the waymo zone and clearly not en route to any maintenance facility either. Stuff like windy canyon roads with no thru access anywhere that are currently gated away, you might see a waymo with a human today.

Waymo is not the only company making lidar maps right now either. I've seen UPS deliver trucks with retrofitted lidar scanners on the roof now. I've even seen this on a police car already, looked like a black rooftop industrial ventilator on a 2ft mast installed directly on the crown victoria roof.

I live minutes from Ford's new HQ in Dearborn and I see multiple lidar equipped vehicles daily. Or at least vehicles loaded up the wazoo.

Waymo does have a database of every today they drive, but for this they don't need one.

If the car comes to a road covered with water, and that road is in the database, and the water level appears low compared to the historical level of the road in the DB, then the car could cross. if the road is not in the DB, then a different decision might be made.

This is similar to humans: you might make different decisions depending on whether you know the road well or not.

Reply on news.ycombinator.com