When Will We See a New Crop Of Autopilots?

[buzzstar]

Auto pilots of the usual present type supplement my sailing. I like sailing solo, and do not want or need a crew, human, mechanical, or electronic. That said, it might be nice to have the latter on board, just so I continue to switch it on or off, and it cannot switch me on or off.

PS #15 has it right I think, but it is even more OK with me since I am 73.

[Hydra]

Ten years ago, there was a research project in the Netherlands for "robot sailing": robosail.nl

It was finished in 2007.

Alain

[Pizzazz]

It is definitely going to happen. Most autopilots hold course better than hand steering (unless you are a racing skipper and 100% focused on the wind ahead). You can still hand steer if you want but it does not hurt to have the capability.

Imagine the possibilities of a truly integrated system: it could steer to a wind angle, calculate the route deviation if the wind changes but still decide not to wake up the skipper to let her have at least three hours of sleep at a time if there are no dangers ahead. Another scenario, achievable today with a little effort: you autoroute to outside the marina, say you need 20 min to get out. You can arrive at the boat and head out immediately, doing all the preparation while the pilot steers you out.

The key challenge in my mind is to get out of the marine electronics realm and feed sensor data to an app that can be developed much faster at a lower cost than what the big manufacturers currently achieve. This means we need a way to interface broadband radar to an app like iNavX or CPN reliably. Also we need custom autopilots (that do not require track change confirmation. Once you get radar and a nice camera interfaced into the PC (along with the typical nmea data), the rest is just software. For a boat you do not need the fancy sensors in the Tesla cars. There is another project coming up on the horizon

Sent from my MotoE2(4G-LTE) using Cruisers Sailing Forum mobile app

[Siberian Sea]

One of the problems with having developed a new " smart " auto pilot that can just about sail your boat for you will be even more dumb sailors at sea.

[Hearts Content]

As stated, conquer downwind ship handling in seas. That would be an achievement.

It's probably the hardest problem to solve as it needs to predict.

[Stumble]

Originally Posted by Hearts Content

As stated, conquer downwind ship handling in seas. That would be an achievement.

It's probably the hardest problem to solve as it needs to predict.

It a problem that has already been solved, really it's been solved for years.

High end autopilots from NKE or B&G (not their retail line) do about 95% of the driving of all long distance solo races. Something like 95% of all mini racing is done on autopilot. Because even in massive waves and big wind the autopilots are better drivers than most humans, certainly after days at sea pushing the boat.

Raymarine/garmin are fine, they are basic reasonable units that handle benign conditions admirably, but they are not even in the same class as high end systems. The processors are faster, they crunch far more data, can use far more sensors, and have much better decision making algorithms.

It's the difference between a Ford Taurus and a Ferrari. Sure they are both cars, but they really aren't the same thing.

[conachair]

Originally Posted by Jd1

The appeal for me is that as a single hander I could sail 24 hrs a day without hallucination from trying to stay awake..

Wind vanes do that

Solo, if you need more control than a wind vane offers then you need to be awake.

[Juho]

There are two possible tasks for an automated navigation system. One is to let you take a nap while on watch, and the other is to not allow you to sleep while on watch. When sailing (in complex environments like archipelagos, as I often do) I like to have a second set of eyes watching the surroundings, charts and meters. The other person may well note something that I may miss.

A good autopilot would do also this - force the pilot to stay alert and not miss signs of danger. We already have some alerts set, like when the water gets too shallow. A good autopilot would make all these warnings more automatic. Instead of just following the chosen direction, it would tell me when it no more makes sense to follow that direction. It could make a big noise and tell me that nearby there is a rock, if I had not already told it that I have taken that rock into account in my navigation plans.

This autopilot would thus extend the current separate warning systems and some additional ones into an integrated system. It would make sure that I will observe everything that is necessary for safe sailing. And it would punish me by giving alerts if it thinks I have missed something. That would be a third navigator onboard, making sure that sailing is safe.

A good autopilot would also make it possible for me to take a nap and wake me up if it observers something risky. It would be my choice if I want to delegate my duties to the automated system, or if I would task it to improve my performance, to keep me awake, to double check that I'm up to date all the time, and to make sure that I will not ignore anything important.

[Sandero]

Seems to me that the actual steeering... turning the helm is a boring and trivial matter. Sensors which evaluate sea state can offer better tracking for sure. However... the intelligent part of getting from here to there is something for the intelligent sailor who processes data from numerous sources and makes decisions such as course to steer. Can a computer process this information faster and more accurately? Probably yes. But giving more "control" to electronics and smart technology leave the skipper with little to do. Heck smart system can even trim sails! Why just let them drive the boat?

Like Kenomac... I rarely helm the boat and usually only getting moving or coming to a stop... so to speak. But I am tweaking the AP's course... I choose waypoints letting the smart system compute course and I program the AP to that course. Not much work there. And the smart system will tell me if I am not making the course and I need to correct for say current.

My involvement is not taxing nor difficult and I just avoid the boring working of being stuck behind the helm. I find I can process the data well enough for my purposes... and so a google type smart AP is not in my future.