When Will We See a New Crop Of Autopilots?

[Jd1]

Tesla has just released their "Autopilot 2.0" with a second camera that can react to traffic lights among other things. Apparently the previous system needed a car in front of it to know when to stop.

So, with this technology starting to go mainstream, I wonder when we will see autopilots for vessels that are 'smart'? What I am envisioning is technology that could use radar, charts, depth sounders, wind sensors, barometers, GPS, weather forecasts and vision systems to all feed to the autopilot.
When cars can drive themselves autonomously through city traffic, I would think that sailing across the oceans should be possible (and much simpler).

Can you imagine telling your boat your destination and maybe some way points and hit the 'go' button before retiring for a nice nap? Floating containers, deadheads, crab traps or whales - not a problem, other ships - not a problem. Even a major shift in weather could be accounted for given the right level of technology.

I have already read stories about freighters/tankers and other commercial ships that will run unmanned or with severely reduced crew. Sailing vessels, although a bit more complicated, should also be able to run on their own. Any situation that the autopilot can not handle such as a major sail adjustment would just result in a notification to the crew (wake up call?).
Who needs crew on watch if technology can do the job?

Obviously there needs to be enough checking going on behind the scenes to know if something is amiss with the system but I am sure that is already taken care of in automotive technology and can easily be adopted. Any kind of failure could easily alert the crew to allow for repair or more likely circumvention of the defective sensor with possibly just a reduced level of functionality.

Exciting times ahead .... !!!!!

Anyway, just throwing this out for discussion to see what people think .....

[muttnik]

Originally Posted by Jd1

Can you imagine telling your boat your destination and maybe some way points and hit the 'go' button before retiring for a nice nap?

No. For the same reason as I don't think Caterham will have much customer demand to licence Tesla's self-driving technology.

[dwedeking2]

At that point just fly to your destination and rent a boat for sitting at anchor. Or better yet just put on your VR glasses and never leave your house. Just make sure to move around periodically so you don't get bed sores...

While an autopilot allows you some freedom of movement around a boat while in transit (your not stuck to the helm), especially when your shorthanded, I think part of the draw of sailing is the "doing".

[Pete7]

Surely the idea of steering is to give the crew something to do on passage after they have swabbed the decks. You don't want them lying around bored that will only lead to mischief. Best keep them gainfully employed.

Whilst the autopilot is selecting the best course to windward for your unmanned yacht it could also adjust the electronic winches and main sheet traveller every time there is a gust so the yacht doesn't lean more than a pre set limit or risk a broach.

Pete

[pickpaul]

Well, Garmin, iNavx and prob some others have autoroute now so you can search for a destination and it plots the course knowing your draft and height. With my Garmin attached to my AP, while motoring it can take me there on it's own but you still have to watch out for traffic. Don't think you'll see fully auto sailboats but definitely in large commercial ships given time.

[Snowpetrel]

I think it will happen sooner rather than latter, they way the big manufacturers like to add bells and whistles.

Technically the problem is easily achievable. Eg opencpn can already easily drive an autopilot, and with ais and radar marpa inputs it wouldn't take much to add autoroute and dodge features.

The biggest question is why you would want to hand over control to a computer. And no matter how good the system is it still wont relieve you from having to keep a lookout.

What I do see happening next is a kind of dodge vessel algorithm using AIS and MARPA that suggests a few safe options for alteration of course to achieve a set CPA from targets.

In heavy traffic this could be very helpful, much like the big ship ARPA's trail maneuver feature, but automated and smarter.

Like a game of chess the algorithm could trail many alterations ahead and help avoid being boxed in. It would need a smart way to deal with traffic altering course. Maybe an alarm and a rerouting feature.

[MarkJ]

So often i butt heads who think sailing boats need paper charts and sextants. These fools living in the dark ages will, no doubt, be agast at Google drone cars safely wandering traffic.
Yes, the will still have accidents... One hit something the other day and the media stoutly reinforced old-timers fears. They will have less accidents, kill fewer etc etc.

Shipping is becoming automated now
Rolls-Royce Drone Ships Challenge $375 Billion Industry: Freight - Bloomberg
The Autonomous Ship | MUNIN



Yes! Its all coming! For big ships and our little ships! Its here now in that every transoceanic record is now set by boats with fewer crew than before, relying on more automation and remote control.

Embrace the spill-over as it will make us safer


Mark

[Kenomac]

We rarely touch the helm, only when anchoring or coming into a marina (which is a very rare occasion). And even then, Auto does half the work.

There's no way I'm going to spend my cruising days sitting behind the helm.

On todays eight hour long crossing, zero time was spent at the helm when I took up the anchor, and less than three minutes spend during anchoring.

[rha9]

I agree that AP advances that allow extensive auto routing over long distances hold very little appeal. What would be a big advance for sailing autopilots at least would be improved performance in more difficult sea states. I know of no sailing AP that does a very good job off the wind in much of a seaway, and conversation with suppliers seems to confirm this. Aside from being uncomfortable, poor performance off the wind could end up making the use off the AP dangerous as well.

[HopCar]

The technology is starting to show up in commercial ships. It's only a matter of time before it works its way down to recreational boaters. It will be up to the individual wether to use it or not.

[leftbrainstuff]

Fully integrated, aspirationally intelligent systems have several limitations:

1) in defence and aerospace circles the loss of situational awareness by users who rely on them results in an increase in incidents and accidents. They fail to recognize enough patterns to be trusted.
2) they have no redundancy due to their tight coupling and therefore are not ideal for mission critical and safety critical systems. They do not degrade gracefully.
3) our ability to engineer optically aware sensors and resolvers is still at a technology resource level of 3. It needs to be 7+ to become ubiquitious.

Sent from my SM-N900T using Cruisers Sailing Forum mobile app

[dwedeking2]

It just seems to me that if you wanted to "take a nap and arrive later" there are better ways (more comfortable, faster, etc) to travel vs a sailboat. To me this means there would be less demand for a "super" autopilot system.

I get for commercial boating as it's more about efficiently moving product (keeping down payroll costs, increased routing efficiency).

[Jd1]

Interesting reactions! I did not see it coming that people would reject the idea just because they like to experience sailing. The thing is, there is an 'off' switch for that!
The appeal for me is that as a single hander I could sail 24 hrs a day without hallucination from trying to stay awake.
I currently find a place to anchor when I get too tired and any crossing that is beyond my ability to stay awake is a no-go. I do not subscribe to the idea of testing my luck by hoping that I am far enough away from any shipping lanes that I can sleep (while sailing or even hove-to).

[Stumble]

Originally Posted by rha9

I agree that AP advances that allow extensive auto routing over long distances hold very little appeal. What would be a big advance for sailing autopilots at least would be improved performance in more difficult sea states. I know of no sailing AP that does a very good job off the wind in much of a seaway, and conversation with suppliers seems to confirm this. Aside from being uncomfortable, poor performance off the wind could end up making the use off the AP dangerous as well.

Current autopilots are more than capable of handeling this. The NKE specifically is perfectly capable of handeling big seas and big winds with a spinnaker up without being babysat. The issue is they are pretty expensive even compared to more mainline ones.

[fallingeggs]

I hear the latest generation of passenger jets hardly even need a pilot these days. And obviously cars are getting close.

What do cars, passenger jets and shipping have in common? They account for trillions dollars in global revenue.

Recreational sailing is a drop in the ocean. Maybe a few billion in sales each year and some small multiple in aftermarket sales/support/services?

Without the cash, we're not going to attract the humans and capital to engineer a boat that will truly sail itself. Helming is only one of the issues. We're just starting to see autopilots that can route a powerboat, let alone a sailboat according to prevailing and predicted whether conditions. Some stand alone apps are coming close to whether routing, but they aren't yet integrated with the autopilot.

And the route is only as good as the boat is trimmed for the conditions. I haven't heard of anything even attempting automatic trimming via a boat's actual polar table and current conditions. The America's Cup boats collect a ton of data that could probably be used to tell the crew just how to trim the boat, but it hasn't happened yet.

Agreed that it is probably well within the technical limits of todays computers to crunch the numbers. (After all, sailing can't be a harder problem to solve then flying, can it?) And might only take small changes to the hardware to effect the trimming of the sails.

Perhaps we'll start to see something approaching full automation in high tech racing experiments or some Russian's $500 million toy, but for us small fry? I predict not before I retire (and I'm 30).