Autopilots and Tides, headings and tracks

[haydude]

I wonder if there are autopilots that can hold the correct heading whatever the course and tide over a long passage and for a long passage I mean at least 8h or more.

Assuming it is calm and that we are motoring for simplicity and thus our course is not influenced by weather, an autopilot can hold a course to either a waypoint or to a heading. A Raymarine autopilot manual states that it has "Tidal Stream Compensation" however the small paragraph does not provide further explanation on how this is achieved. It states too that the autopilot will "maintain track". I thus assume that compensation is all about keeping the boat on the track line that joins the point of departure to the arrival waypoint. This might work well for a short hop of less than one hour, but I am not sure it works with longer passages.

We assume that we are in the English Channel and tides move East/West (for simplicity).

On a course East to West the influence of tide on my course will be assumed as negligible and thus I will navigate by setting a waypoint to destination (assuming a straight course with only the destination waypoint).

However if I sail North to South (for example on a passage to France), if I set the destination as a waypoint I believe the autopilot will try to maintain the boat on the same track compensating over the 10 to 12 hours passage time, thus making my passage longer.

In this case I prefer to calculate a heading to hold, allowing the boat to "swing" its course. For example if I head South at low tide, my boat will swing East for the first six hours and West for the following six hours. Let's say my course destination is 60 nautical miles away at 180 degrees heading, and my speed is 5 kts, my heading on the Autopilot will be bang on 180 degrees.

Given that charts now hold also tidal stream data, an autopilot could use it to calculate the correct heading to maintain (rather than a track), allowing the boat to swing temporarily off track as I do when I set my heading.

The question is: are there any autopilots that can do this?

 

[bbg]

I wonder if there are autopilots that can hold the correct heading whatever the course and tide over a long passage and for a long passage I mean at least 8h or more.

Assuming it is calm and that we are motoring for simplicity and thus our course is not influenced by weather, an autopilot can hold a course to either a waypoint or to a heading. A Raymarine autopilot manual states that it has "Tidal Stream Compensation" however the small paragraph does not provide further explanation on how this is achieved. It states too that the autopilot will "maintain track". I thus assume that compensation is all about keeping the boat on the track line that joins the point of departure to the arrival waypoint. This might work well for a short hop of less than one hour, but I am not sure it works with longer passages.

We assume that we are in the English Channel and tides move East/West (for simplicity).

On a course East to West the influence of tide on my course will be assumed as negligible and thus I will navigate by setting a waypoint to destination (assuming a straight course with only the destination waypoint).

However if I sail North to South (for example on a passage to France), if I set the destination as a waypoint I believe the autopilot will try to maintain the boat on the same track compensating over the 10 to 12 hours passage time, thus making my passage longer.

In this case I prefer to calculate a heading to hold, allowing the boat to "swing" its course. For example if I head South at low tide, my boat will swing East for the first six hours and West for the following six hours. Let's say my course destination is 60 nautical miles away at 180 degrees heading, and my speed is 5 kts, my heading on the Autopilot will be bang on 180 degrees.

Given that charts now hold also tidal stream data, an autopilot could use it to calculate the correct heading to maintain (rather than a track), allowing the boat to swing temporarily off track as I do when I set my heading.

The question is: are there any autopilots that can do this?

I was able to set my autopilot to steer to a compass course, a waypoint, apparent wind angle and true wind angle. I expect that the first thing most autopilots steer to is a (fluxgate) compass course. Then you can add the other possibilities as you integrate other instruments.

 

[prv]

A Raymarine pilot steering to a waypoint will draw an imaginary line from your current position to the waypoint at the moment you engage the pilot, and then it will do its best to keep the boat on that line across the ground.

You can (and I do) work out a course to steer allowing for tide, and then tell the autopilot to steer that course, so that you follow the efficient C- or S-curved track. You cannot, with traditional Raymarine stuff, have the system compute the required course all on its own.

I'm not aware of an autopilot per se that can do this, but I believe there is some PC navigation software that will work out the course. It's just possible that the software can output NMEA data to drive an autopilot appropriately, although if I had to bet I'd say it probably doesn't.

Pete

 

[Giblets]

Does the autopilot not have to be connected to a suitable plotter for this to be activated. Our Raymarine ST1000+ can do this but it has to be connected to a suitable navigation system operating in Track mode which we haven't tried yet.

From the manual:

2.4 Using Track mode
In Track mode, the tiller pilot maintains a track between waypoints created on a
navigation system. The tiller pilot computes any course changes to keep your boat
on track, automatically compensating for tidal streams and leeway. To operate in
Track mode, the tiller pilot must receive cross track error information from either:
• a SeaTalk navigator, or
• a non-SeaTalk navigation system transmitting NMEA 0180 or 0183 data.

 

[Tranona]

Does the autopilot not have to be connected to a suitable plotter for this to be activated. Our Raymarine ST1000+ can do this but it has to be connected to a suitable navigation system operating in Track mode which we haven't tried yet.

From the manual:

2.4 Using Track mode
In Track mode, the tiller pilot maintains a track between waypoints created on a
navigation system. The tiller pilot computes any course changes to keep your boat
on track, automatically compensating for tidal streams and leeway. To operate in
Track mode, the tiller pilot must receive cross track error information from either:
• a SeaTalk navigator, or
• a non-SeaTalk navigation system transmitting NMEA 0180 or 0183 data.

However, it is not "automatically compensating" based on live tidal data or leeway (which is always a bit of guesswork anyway), but changing course to minimise cross track error from the chosen waypoint. End result is that the aP is always working to alter course if there is strong tide or variable leeway to always head for the waypoint. OK if you are in tidefree waters like much of the med and particularly under motor, but not helpful in our waters under sail.

Sailing to the apparent wind is useful, but for most purposes the method described by Pete is preferred - plot a course the old fashioned way and steer to a heading, unless the resulting course over the ground puts you in danger. you still might have to deal with wind shifts, just as when steering manually.

 

[laika]

Does the autopilot not have to be connected to a suitable plotter for this to be activated. Our Raymarine ST1000+ can do this but it has to be connected to a suitable navigation system operating in Track mode which we haven't tried yet.

I believe this is steering to a waypoint: the compensation for tidal streams is just changing the heading as and when the stream pushes you in a particular direction resulting in a longer passage on a cross channel [EDIT: As Tranona beat me to saying: minimising XTE]. Haydude is (I believe) talking about software which can do automatically what we currently do manually, i.e. calculate a singe course to steer for the entire taking into account currents over the course of the distance to the waypoint to minimise distance travelled in the water. As prv says, this is not going to be a function of the AP per say but whatever controls it, i.e. a laptop or plotter. My C90W certainly can't do this but it has come up before and I *think* someone mentioned something (Explorer? Maxsea? don't remember) which would do such calculations but really don't remember.

 

[Giblets]

Neptune Planner Plus calculates a course based on predicted tidal streams throughout the trip along with boat speed & predicted leeway but I don't think it can be fed to a AP. Once such a course is obtained surely the AP can be set to the calculated course.

 

[Twister_Ken]

Sounds like the OP might be looking for too much electronic input. Surely better to do the sums yourself, and then punch a CTS into George. Even if George MkII could do it all himself, I'd still want to do the sums to make sure that he was right!

 

[bbg]

Sounds like the OP might be looking for too much electronic input. Surely better to do the sums yourself, and then punch a CTS into George. Even if George MkII could do it all himself, I'd still want to do the sums to make sure that he was right!

That was my thinking as well. BTW, the autopilot is not George, it is Britney.

 

[Giblets]

That was my thinking as well. BTW, the autopilot is not George, it is Britney.

Hissing Sid here!

 

[bbg]

Always Britney. Britney Steers.

 

[prv]

Sounds like the OP might be looking for too much electronic input. Surely better to do the sums yourself, and then punch a CTS into George. Even if George MkII could do it all himself, I'd still want to do the sums to make sure that he was right!

I'd be willing to trust a machine, on the grounds that being used to doing it myself I'd probably notice if the result was grossly wrong. I don't have any great wish to acquire such a machine though, the chore is hardly onerous.

Pete

 

[prv]

Always Britney. Britney Steers.

Groan.

Ours is George, I believe it's an RAF thing.

Although my mate's 5-year-old nephew apparently thought it was the man from Southampton VTS doing the steering . Presumably because his voice was emanating from the binnacle next to the wheel.

(This is of course where the remote handset for the VHF is )

Pete

 

[laika]

Sounds like the OP might be looking for too much electronic input. Surely better to do the sums yourself, and then punch a CTS into George. Even if George MkII could do it all himself, I'd still want to do the sums to make sure that he was right!

Not sure. I'm good at sums but trust a calculator more. My cross-chanel plotting method involves transferring vectors from the tidal atlas to a piece of paper fairly freestyle which isn't very accurate. Yes I'd want to do the calculations to cross-check the first few times but if the software was consistently as good or better at it than me I might be tempted to use such a system if I was already a fan of electronic navigation. Besides, you normally have an idea *roughly* what the course should be so you can roughly sanity check it. Combined with wind predictions and boat polars a computer could doubtless do a better job than me. Moreover if hooked up to the AP, software could presumably dynamically update the optimum heading based on how wind/tide matches what was predicted.

Note the "if I was already a fan of electric navigation" caveat. Personally I'm not. But I see the value for those that are.

 

[jerrytug]

That was my thinking as well. BTW, the autopilot is not George, it is Britney.

Please tell me that's not Britney Steers

 

[prv]

My cross-chanel plotting method involves transferring vectors from the tidal atlas to a piece of paper fairly freestyle which isn't very accurate.

And indeed I'm sure the underlying surveyed tidal data for the Channel is far more accurate than what's printed at relatively small scale and wide spacing in the atlas, even before I inaccurately combine it with my approximately predicted position.

Pete

 

[haydude]

Not sure. I'm good at sums but trust a calculator more. My cross-chanel plotting method involves transferring vectors from the tidal atlas to a piece of paper fairly freestyle which isn't very accurate. Yes I'd want to do the calculations to cross-check the first few times but if the software was consistently as good or better at it than me I might be tempted to use such a system if I was already a fan of electronic navigation. Besides, you normally have an idea *roughly* what the course should be so you can roughly sanity check it. Combined with wind predictions and boat polars a computer could doubtless do a better job than me. Moreover if hooked up to the AP, software could presumably dynamically update the optimum heading based on how wind/tide matches what was predicted.

Precisely. I trust that my Navionics charts with tides and tidal flows are more up to date and precise than my tidal atlas that summarizes most of the East of West end of the channel in one page. The information is there, only it is not being used to calculate a course. So my E80 Classic is no worse than the newest plotter+pilot combination (at least regarding this capability or lack of).

There might be a gap in the market for a new product then.

The value added of an electronic calculation is that it can be re-calculated real time depending on condition. When I do manual calculations I allow a worst and average case scenario and use that for the passage. Whilst underway I no longer redo these calculations, I just compensate for error when I am nearly there. In general the error is small, however on one occasion I allowed 11 hours for a return passage South to North, a favorable wind could have taken me to destination in 8 hours flat, but I had to sail an additional 40 minutes to compensate for the error because I did not get the whole of the ebb tide near destination.

That is a situation when a computer could have recalculated the correct heading (course to steer) looking at the real average cruising speed whilst under way.

 

[RobbieW]

Given that the tide strength & direction on charts, whether paper or electronic, is not absolute it must be subject to error. You really dont know the error until you are doing the passage as it will be different for each one. What I do is use a proforma with HW -6 to +6 for two 12hr periods, so I can either use a day of one port or two half days for different ports, then copy across position based estimates of tidal direction/strength much as Laika described and do a rough vector sum to derive CTS. I then have a record of my estimates which can be used to, fairly easily, correct the course as necessary when the tide changes - always bearing in mind the desirability of being up tide of the destination in the last hour or so!

 

[Twister_Ken]

With an all-singing all-dancing George, you'd also have to have a word in his ear about leeway.

 

[lpdsn]

There might be a gap in the market for a new product then.

And once you have that you can go for the version of the autopilot that actually turns up in the classroom and sits your YM shorebased exam for you.

On a serious note, I can see the advantages of it, but I'd be wary of ending up like those Air France pilots that had forgotten how to fly.