Automating Road Maintenance with Lidar Technology | Hacker News Reader

Automating Road Maintenance with Lidar Technology

spectrum.ieee.org

It's not like we don't know where the potholes are, it just costs too much to maintain all the roads. Spending more on Lidar cameras can't fix that.

Do we really know where potholes are though, I would think not. I mean when you drive over it you know, hey pothole here, and probably people know hey this stretch of road is badly maintained but unless you have an accurate map of where each pothole is then it follows what you do is to select a stretch of road and say we will go down this stretch and fix it this month.

Which would seem inefficient if you had a cheap way of amassing exact data on where precisely all the potholes actually are.

So when you drive over that pot hole, as a responsible citizen you call the local highways department and report it, then they know where it is. At least that's how it should work. Fairly cheap and effective.

>So when you drive over that pot hole, as a responsible citizen you call the local highways department and report it,

ok, so I guess I have never, ever, ridden with a responsible citizen.

I have ridden with people who, on the way to important meetings, see cars that look abandoned out in a field and they get out and investigate to make sure they really are abandoned, but not responsible enough to call up 'somebody' and report the locations of potholes as encountered?

In Europe it's normal; if you encounter a road safety issue you report it and a crew gets sent out to fix it. I think in the US people stopped giving a shit, because you know it probably won't get fixed so any time soon so why bother?

Never heard about such things. There are public servants which inspect the state of the road.

"There's an app for that." I mean, right ?

In Toronto, Canada there is:

* https://www.toronto.ca/home/311-toronto-at-your-service/311-...

A typically Canadian solution. Communitarian and very Bolshevik :-P

Well damage caused by potholes is exponential. So efficiently repairing various kinds of potholes with various technologies requires knowing where and what type the potholes are.

Ideally all potholes would be repaired when tiny, and easy to fix.

I wonder if it's actually more cost effective to let them get to a size just before they actually start breaking cars. I have 0 knowledge of road maintenance but I reckon getting people, material and tooling to the pothole would be the bulk of the cost.

Well the problem is the damage goes exponential. Roads degrade with pressure changes and friction. But cracks let water in, which can result in erosion (of the road or underneath that supports the road), and then any elevation changes can greatly increase wear (for instance a 50,000 pound truck going 50 mph is regularly hitting the far side of the pothole).

Even a single winter can make a dramatic change to average road quality, thus the hurry every summer to repair roads.

This is a great idea. I'm shocked no one (to my knowledge) has tried this before. And I mean that in the best possible way, which is to say that great ideas seem obvious in retrospect.

I just hope it's largely installed in electric vehicles if they're going to drive around most of the time.

This is essentially a commercial service already: https://www.mobileye.com/en/data/

A single GPS receiver might not be accurate enough. Differential GPS could help here.

You don't need better than +/- 50 feet.

That's enough to get you on the correct street.

The bot doing the repair can figure out exactly where the pothole is.

Seems like a good application for accelerometers as well.

Which means that both Apple and Google already have the data.

Reply on news.ycombinator.com