Originally published at: https://boingboing.net/2018/05/30/autonomous-boats-in-some-citie.html
…
Imagine
You’ll never guess what I’m doing right now?
Probably not but do you mind if I join in?
MIT to Venice/Amsterdam: “You know, I have a wonderful idea. You should use autonomous boats! Smart eh?” Amsterdam and Venice: “Yeah, well… Thanks for the tip man.”
Autonomous boats might be a good idea in cities with largely unused waterways, but in Amsterdam/Venice they are probably even harder than autonomous cars. At least a car doesn’t need to compensate for wind drift.
Nice idea. Not the best example.
[edit:]
Not to say that autonomous boats won’t be useful in the (near?) future, they’re not that much harder than cars after all. I just find it a bit silly they propose to ease congestion by moving to the water. Have they even been to those cities I guess. Especially in Venice the water is as congested as the land.
Things autonomous cars have to deal with: fast moving (>25mph) vehicles, significant numbers of pedestrians, bicyclists (and people on skateboards, scooters, wheelchairs, etc.), random animals, complex road signs, road markings and lighting systems, changing road conditions due to road work and weather conditions, etc. etc. In comparison, wind drift seems like something that’s measurable and can be compensated for, algorithmically, without too much trouble. Everything else an autonomous boat has to deal with has an analogy with autonomous cars. So I’d say that autonomous boats are significantly easier than autonomous cars - the speed issue alone would make that so (I mean, the maximum speed limit for a boat in Venice is about half the minimum speed limit for a car in the US).
Do You Want Transformers? /Archer
Tell me more of such cities, for I know none.
I’ll tell you what, though. Autonomous boats are an awesome idea for automating the process of cutting my fishing line like a douchebag.
Maybe you are right. the slower speed will make failures at least less disastrous, probably. Though the slower speed is balanced by a larger mass in many cases.
Still, I don’t think autonomous boats will be easier than cars. Computers are quite good at speed. But not that good at recognizing objects. And my guess would be that water with lots of waves and wind would be a harder environment for computer to deal with than a (relatively) smooth street.
The waters around Manhattan are remarkably underused, especially compared to a hundred years ago. I kayak there, keep away from the few ferry lines and there’s really not much at all.
We’re talking about something between 1/8 and 1/20 the speed of a car. Venice canal speed limits are less than 4mph - slow moving vehicles give so much more time to identify objects (both stationary and equally slow moving) and figure out how to respond to them. It’s not the speed of the vehicle itself that the computer has a hard time with; the autonomous vehicles of the early '90s were traveling at 5mph because of the processing time it took to identify what was going on in the environment and respond appropriately. Computers still aren’t that great at identifying objects - which was part of my point - there are a lot more objects needing to be identified on roads and a lot less time in which to do so. Wind and currents (which are less of an issue in canals) can be detected and calculations done fairly easily to adjust for them. Again, cars have similar, but more complex issues - cars, too, have to deal with wind sometimes, in addition to road conditions that impact how wheels turn and handling works (e.g. loose gravel, water, ice).
Vehicles on roads have a lot more issues to deal with and a lot less time in which to do so. That’s simply more difficult. That’s not to say that either is easy, though.
This topic was automatically closed after 5 days. New replies are no longer allowed.