▲ 77 points
back
28 comments
Same news but from other source: https://news.ycombinator.com/item?id=29024204
Disclaimer: I'm the founder of blkSAIL (black sail) that builds AI and autonomy for ocean-going maritime ships. We are an MIT CSAIL spinoff. We started with an aim to become Uber for the waterways, then pivoted to large ships.
Many of the comments below touch on many key points: (1) most commercial ships have a crew between 8 to 20, (2) auto-pilot in ships exist since the GPS was invented, (3) when ship hits the fan, there's not much you can do.
(1) Autonomy in maritime is not to replace seafarers, the navigation crew is less than 30% of the crew. And crew is negligible OPEX compared to fuel etc. (2) Because the industry is the backbone of global trade, it moves much slower in adopting technology for the sake of technology. However, when there's a simple reliable system that can clear benefit, the adoption happens. So given that there's few[1] lanes in the ocean, you can easily go from one waypoint to the next. (3) Onboard a ship, there's no such thing as breaks. In most cases, you know about a collision or a grounding 12+ minutes in advance. The challenge is in getting the predictions right and abide by the rules of the road. Most advanced LIDAR have 1km range, which is too close, cameras don't see much neither[2].
[1]: there are quite few channels around ports to ease traffic that are one-way. [2]: the resolution is too low to see far out. When using zoom, the stabilization is a nightmare.
> (1) Autonomy in maritime is not to replace seafarers, the navigation crew is less than 30% of the crew.
As someone who works in the autonomous ship space now and a prior merchant mariner, I agree with a lot of what you state, except for point 1...
In the near term, yes, I agree that people are working to use autonomy to augment the bridge crew (and engine crew and ship owners for that matter). However, in the long run (and there are actually several prototype autonomous vessels on the water today), the ultimate goal is absolutely to eliminate the need for a human crew.
Reasons... The biggest driver is that much of the design of ships is driven by the fact that human crews are embarked. There are too many items to list, but human factors include the ship superstructure (living qarters/bridge/galley/sanitation), freshwater generation, HVAC (a large chunk of energy costs, safety provisions, medical, and on and on. Other drivers are 1.) Human safety, 2.) Asset (ship & cargo) safety, 3.) Litigation avoidance/lower crew/cargo insurance/underwriting - which relate to 1 and 2 (e.g. collisions into other ships, piers, or getting your ship wedged in a canal). Many maritime incidents are the result of human error.
To the naiive, it may seem that autonomous operation at sea is easier than driving autonomously on land. However, the lay person may not realize that ships come in a variety of shaps/sizes and maneuvering abilities (for different functions including ferries, tankers, containerships, taxis, navy vessels, barges, cruise ships, etc.), communicate via complex bridge-to-bridge communications, navigate using a variety of floating/light/radio indicators, and are very sensitive from a stability and structural soundness perspective to weather (seastate, wind, currents, etc), among many other complex constraints. Thanks to inertia and the large absence of "traction", large ships don't stop instantaneously, they have a stopping distance which can be over 4 nautical miles [1] !
Driving cars autonomously is very difficult, but driving ships autonomously is just as difficult, if not more so.
1. https://knowledgeofsea.com/stopping-distance-turning-circle-...
Wow, I can't believe this is happening only now. This is such a great idea. Autonomous boat driving is likely a lot simpler than car driving. Mostly because there are only few other boats, each with their own inertia, no roadwork, little tunnels, no pedestrians... This is definitely part of the future...
> no pedestrians
And swimmers!
My office here in Amsterdam is on a canal. All through the summer there are 20-30 swimmers a day. Including one man who every morning swims head-down freestyle the 1km around Prinseneiland.
As someone with chronic sinusitis, which likely came from years of swimming in questionable water, this is utter insanity to me.
Does canal swimming seem a bit nuts to you as well?
We keep our canals reasonably clean; especially if you're close to one of the inlets of clean water the risks are low from what I understand. In the summer you have to watch out for algae etc. But we also keep track of those in official swim water around the canals. And finally I think the risks of swimming in a crowded small official swimming lake can be pretty high as well. But yeah, I'd still think twice before going swimming in a canal and I'd try to keep the water out of my mouth etc.
> no pedestrians.
Actually there are plenty of Single sculls, kayakers, paddleboarders, etc. in most canals in the Netherlands.
But indeed the reaction time you need is still a lot slower than with pedestrians jumping in front of a car.
It is not as simple as that. Unlike roads which are mostly deadlock free (of course with some exceptions), going around in Amsterdam canals can be extrmely tricky. There are big tourist boats that are way too big for the canals, so you have to move out of the way when they turn. There are plenty of bridges where only one boat at a time can pass. And there is no requirement to keep right on the water.
Going on bigger water is also not easy. Ferries accelerate very quickly when they depart and have priority over small boats.
For bigger boats, you have to assume they cannot see a smaller boat and if they see one, they cannot brake or change course.
There are already some companies experimenting with autonomous ships in international waters. Maritime drones have also been a thing for a while. There's a bay area company that has been operating some of those for a few years. They are at sea for months.
And most evidence for failure will quickly remove itself from the surface of the earth. /s
Looking forward to autonomous driving on rails /s
Jokes asides, I feel like autonomous driving might technically be made to work by allowing it only on certified roads, with standardized conditions. At that point we might have reinvented trains.
I think it all pretty much boils down to:
- Roads are cheaper to lay and maintain than rails
- Roads are infinitely more flexible. You cannot walk, bicycle, or hold a parade on a railway track.
I agree with the flexibility point 100%.
On the cheaper I really do not know. Is pouring asphalt cheaper than putting rails? I honestly do not know.
It also might depend on the implementation. I remember a history teacher mentioning that the uk built railwways providing time-limited exploitation licenses for private companies that built them (after x years they became public) and they built a vast network quickly . ( I found this pdf: https://www.campop.geog.cam.ac.uk/research/projects/transpor... about the expansion, but it does not mention the licenses)
In the car case we need to also consider the "wagon", and the fuel cost is outsourced to the customers.
Cannot bicycle on a railway track? Danny MacAskill begs to differ.
Who said we need rails! All we need are roads that are closed to human drivers.
Cool concept but the name doesn't work so well when spoken, as it conveys something considerably lower tech.
I'm assuming it's supposed to be clever (and it is), but still.... it's gotta get tiring always having to clarify that there isn't a W.
Excellent work! But I’m biased having worked at DRL and senseable.
Maybe that's why the LIO-SAM authors kept the Amsterdam dataset private ;) Congrats nevertheless!
Good luck to these robots on the next king day.
There wasn't a Dutch university available?
I think the problem was there wasn't a US canal available.
Seriously: this shows to me that people will use any infrastructure that is most convenient/fast for them. If there are highways people will drive. Build public transport and people will use that. Build bike paths and people will bike. If you have canals like Amsterdam, well...
Canals are probably not very fast nor scalable for mass transit, but they may be useful for transporting heavy things.
This is intended as an autonomous taxi, designed for a sprawling urban canal system like Amsterdam. All those canals you've linked are absolutely nothing like the intended definition.
The ones I linked are for transportation, you could make autonomous ships and test them in canals for cargo, doesn't have to be for passengers. GP said there aren't canals in America, was simply pointing out there are.
Was wondering the same. Originally some were involved according to this 2016 article: https://www.technologyreview.com/2016/09/19/107644/fleets-of...
I bet there are several Dutch on the team who decided to go to MIT instead of their local University.
aaron swartz would be proud