Why auto routing is a bad idea

[Jklotz]

Just today, I dipped my toes into auto routing, as we head south for the winter. Here were 2 places that would have put me on the BoatUS shame list, had I not been present and intervened. In the first example, had my draft been less than one foot, we could have saved a few minutes cutting the channel. I draft 4.5'. Could have made for a bad day! The second example, well, unless my boat learned how to fly, well, you get the picture.

After my 1st experience, trying different platforms, and entering my boats critical info preciously in both beforehand, I'm on the fence. On one hand, the routes populated in seconds. OTOH, I spent 9 hours perched over my plotters, watching like a hawk during my passage today. Good thing I did! I guess the question that one should ask is, is it worth using auto routing then meticulously scrutinizing every way point, editing, then spend the next day wondering if you might have missed something and walking on eggshells, or just to create your own routes from scratch to begin with and forego the uncertainty?

We head out in a few days. Maybe between now and then, I'll chart a course both ways and time my efforts. I suppose that's the only way I'll know if auto routing is a time saver or a stress inducer.

 

[mvweebles]

Just today, I dipped my toes into auto routing, as we head south for the winter. Here were 2 places that would have put me on the BoatUS shame list, had I not been present and intervened. In the first example, had my draft been less than one foot, we could have saved a few minutes cutting the channel. I draft 4.5'. Could have made for a bad day! The second example, well, unless my boat learned how to fly, well, you get the picture.
View attachment 169232
View attachment 169234

After my 1st experience, trying different platforms, and entering my boats critical info preciously in both beforehand, I'm on the fence. On one hand, the routes populated in seconds. OTOH, I spent 9 hours perched over my plotters, watching like a hawk during my passage today. Good thing I did! I guess the question that one should ask is, is it worth using auto routing then meticulously scrutinizing every way point, editing, then spend the next day wondering if you might have missed something and walking on eggshells, or just to create your own routes from scratch to begin with and forego the uncertainty?

We head out in a few days. Maybe between now and then, I'll chart a course both ways and time my efforts. I suppose that's the only way I'll know if auto routing is a time saver or a stress inducer.

There's a saying in business: "you can't outsource risk " Meaning no whining about stuff that was supposed to do tasks that are your responsibility. There's a place for auto-routing. Plug-and-play isn't one of them.

Good examples.

Peter

 

[SteveK]

Chart plotters take the shortest route based on your settings which include draft. I enter 10 feet which keeps me away from shallows. What number did you enter?

 

[Jklotz]

Chart plotters take the shortest route based on your settings which include draft. I enter 10 feet which keeps me away from shallows. What number did you enter?

I just put in my draft, 4.5' and added 3' as a buffer.

 

[SteveK]

I just put in my draft, 4.5' and added 3' as a buffer.

Then it should not have taken you over less. I saw your depth 45 feet and see that your shortcut may have worked if you had set depth 4.5 foot.
I used to moor up river and there was a shortcut over the mud flats that we took at a certain tide level while it was rising, so been there. That was local knowledge. I would stick with your settings for the auto into not often used travel areas.
BTW, I glance at the chart plotter regularly. Mostly because GPS could be off quite a bit. Last summer I crossed over (on chart) a marked rock, which was not under the boat, so where did it go, to the left, to the right.

 

[North Baltic sea]

I have obviously been lucky, as I always use automatic routing. Sometimes it suggests a longer route than necessary, and I don't follow it, but it has never suggested anything dangerous.

I use the Navicons program on my tablet, and my chart plotter is so old that it doesn't have any automatic features, but it does have the Navicons map.

 

[Iggy]

I do use Auto Routing to start. I does about 80 to 95% of the work for me. But I do expect things to be wrong and than I change them. I don't think there is one time that I never made a change.

 

[CaptTom]

I've used it to rough-in a general route, but I always verify and usually refine it based on local conditions and sometimes outright errors it makes.

What I don't understand is how it can be considered a waste of time to make sure your route is safe and to maintain a proper lookout the whole time while following it.

 

[Portage_Bay]

I don't use auto routing much. Two reasons. The first being I've seen too many errors. Well, not errors but making decisions about the route that I disagree with. And generally far too many waypoints. The second reason is that I enjoy creating my own routes. Studying the chart, making my own decisions based on all factors. Even adding distance to take a more interesting route.

 

[drewread]

I haven't used a single piece of autorouting software that doesn't ask you to check and verify the navigational route. You absolutely must check the route.

That said, it is a lot easier to fix a few places along a route than it is to create the entire route from scratch.

If you find the software repeatedly making errors - contact the support department with some screenshots and give them a chance to make it better.

 

[PierreR]

I use auto routing regularly and thought about what is said here and realized that I don't set the auto pilot to follow it, I am always on the bridge to adjust the pilot and always read the water visually. I never trust being exactly where the chart plotter says I am and always check the ATONs I have not run aground yet.

 

[Jklotz]

What I don't understand is how it can be considered a waste of time to make sure your route is safe and to maintain a proper lookout the whole time while following it.

I must have not represented it clearly. The question is not proper route creation and monitoring being a waste of time, it's weather or not auto routing is a waste of time.

 

[mvweebles]

I enjoy creating my own routes. Studying the chart, making my own decisions based on all factors. Even adding distance to take a more interesting route.

Me too - I enjoy passage planning. I find the process interesting and informative.

The closest thing to auto-routing I use with any regularity is PredictWind or FastSeas weather routing that returns a route that minimizes adverse conditions. But that's really mostly useful for open water crossings.

BTW - for those heading to the Bahamas, FastSeas offers route planning based on ocean currents (you have to pay $499/yr with PredictWins). FastSeas is around $60/yr, or you can get three free route plans per month. Caveat - ocean current models can only be developed a few days ahead. No 14-days available.

Peter

 

[DDW]

Auto routing on chart plotters seems to be a solution presented to lower the workload of creating a route. But chartplotters (every one I've used anyway) have horrible interfaces for creating routes. Some laptop software - OpenCpn even - are far better and much quicker. And you get what you want. I'd have preferred the programming time put into autorouting algorithms were instead put into improving the UI on generating a route in the first place. It is a solution to a problem of their own making. Editing a route on chartplotters is just about as bad, so editing their poorly drawn route often takes longer than creating a proper one.

I never use autorouting because it doesn't do what I'd do and usually not close. I use OpenCpn, a days route of 20 or 30 waypoints takes about 10 minutes to create if intricate, is exactly what I want, and gets transferred to the chartplotter (another 3 or 4 minutes). And I am now familiar with everything along the way. I've got 6 hours of motoring ahead, what else do I have to do?

 

[Moonfish]

My primary concern with auto routing is not with the average TFer that has experience and common sense, but with the newer boater that doesn't have the requisite knowledge to understand when they may be putting their boat and crew in danger.

 

[ABfish]

I trust autorouting as much as I trust the current state of artificial intelligence. I'll ask the question and perform due diligence on the answer until I'm comfortable.

I've tried downloading routes from my tablet to the chartplotter, but my favorite method is to build the route directly on the chartplotter using a handheld remote and a paper chart (if available).

 

[MYTraveler]

Fishing for tuna out of southern California often entails a late night departure and travel from 80 to 120+ nm offshore to arrive just before dawn at the "spot" (ie, the offshore bank at which the most fish were caught the preceding day). Those who leave earlier can travel at a more leisurely pace, while those who are willing to travel faster (sometimes more than 20 knots, even on a moonless night), can leave later. The vast majority of boats set a course to the spot and travel in a straight line. Since most boats leave from Point Loma, the vast majority are on the exact some route. After having been passed within feet by boats traveling at a high speed with apparently no one at the helm, I always set a course that would take me about 5 miles from the course that most everyone else was on. I think the same phenomenon might happen with auto routing.

 

[slowgoesit]

We use autorouting all the time, but verify that it is doing what we want it to do . . . . then we don't couple the autopilot to the route unless we are in open ocean. By that I mean that we manually change course to keep up with the route displayed on the chartplotter. That way it keeps us more involved in the course, as well as more involved looking for debris on the water, as well as other boats, ATON's that are out of position, etc. We're not the most experienced boaters out there by far, but it's worked for us over the last 10,000 miles. YMMV

 

[ABfish]

@MYTraveler I encountered a large Nordhavn recently who was hugging the magenta line on the ICW. The captain called ahead to tell me that the area was "very tight" and that he would not deviate from his course. In other words, I should give way regardless of the nav rules.

The area is in my home waters and is not "tight", with plenty of depth and width in the channel. I chose not to argue with him.

I like your thoughts on steering off of the most direct route between popular destinations.

 

[sjisailor]

I use auto routing to get a general plan, but then manually set heading (or directly steer). Generally I parallel the auto course but not always.

I *never* engage the AP to automatically follow the displayed auto course. Many reasons: lack of trust, the need to be vigilant, the fact that the risk comes down on me, occasional observations of weird routing or other glitches, and the fact that other vessels are often following the exact same AP courses ... coming head-on!

And right, some folks out there don't understand this. A friend with a new boat told me that he was planning to use autopilot so he could take Zoom meetings in the salon while underway (!!) Since then I assume no one will see me and I maintain a better watch!

(BTW plot-and-parallel is similar to what is called "slop" in aviation: "strategic lateral offset procedure" to parallel a charted airway but with some miles of offset from it. This reduces risk vs. being on the exact same airway with everyone else.)

 

[Jklotz]

A friend with a new boat told me that he was planning to use autopilot so he could take Zoom meetings in the salon while underway (!!) Since then I assume no one will see me and I maintain a better watch!

Pretty sure I passed your buddy on the ICW yesterday.

 

[Seevee]

​

Jkolts; Whose autorouting do you use?

I love autorouting and use it all the time. I never check it, just plot and go. I do hook it up to the AP.

The only thing it does is steer the boat and make my life easier to do other necessary chores.

There's no guarantee that manual plotting is any better and a lot of the hazards are not known regardless of how you plot.

While I find Simrad more exact, Garmin is easier to use and does longer routes.

I'll monitor the movement of the boat no matter whose steering, and find it easy to go into heading hold or manual to steer around objects and hazards.

Also, I'm an intercoastal cruiser and use it in rather tight areas, too. Works fine.


And the two examples that Jkolts would be ease to work around, if he was looking at the screen, which he obviously was.

 

[SteveK]

There is auto routing on a chart plotter and there is autopilot that is then connected to the chart plotter to receive commands for course corrections.
If it is the former than it is for manual guidance, manual input to the autopilot, but if for the latter, all auto, that can be scary if you allow it to take you along a course without supervision.

 

[Mambo42]

I think I have used auto-routing maybe once and was not impressed. An enormous amount of way points and a route that I did not want to do. So I never used it again and I just keep planning my own routes. It may take a bit longer, but at least I know I will have the route I want to take, plus I have seen the entire route, I took conscious decisions why I placed a waypoint at a certain position, so the route is already in my head. Perhaps that comes from my flying days, but it still works for me.

 

[mvweebles]

At least some of my hesitation is I don't fully trust charts. Maybe I have trust issues, but where does the depth information come from? Crowdsource sounds interesting, but how do you know that folks have their sounder configured properly (or even the same - water depth vs below-keel depth)?

"Trust but verify" comes to mind.

Peter

 

[Frosty]

Crowdsource sounds interesting, but how do you know that folks have their sounder configured properly (or even the same - water depth vs below-keel depth)?

I was thinking about that this summer. I'm just guessing but perhaps they don't "count" the input unless many boats offer up the same data?

That said, I've noticed that many boats set their sounder's Zero point at the bottom of the keel. That drives me nuts (just my opinion). Even watching their videos, they are telling the camera something like "Ooh shallow, we're in 5'. I mean, that's 5' below the keel, and our keel is 3.5' so that's really 8.5' of water." Argghhhh, just set the sounder so it measures from the surface! That way it matches the chart and no need to explain. We all know your boat doesn't draft zero, and with actual depth from the surface can adjust our thinking automatically for your boat, our boat, a friend's boat, or whatever.

But so anyway, yeah, what if 20 of those boats send in a 5' data point, but it's really 8.5'..... so I STILL wonder how it works.

Edited to add: I know on my MFD you tell it where the transducer is, whether you want the depth to show at that depth, at another depth (surface) or at yet another depth (bottom of keel). Maybe somehow whomever receives this "crowd sourced" data gets that info too? Mysterious.

 

[SteveK]

Chart depth is based on normal low mean tide? I actually set my plotter to about 3 feet below my keel. Gives me time to react when the alarm sounds. Auto routing keeps me away from that depth.

 

[Jklotz]

I think I have used auto-routing maybe once and was not impressed. An enormous amount of way points and a route that I did not want to do. So I never used it again and I just keep planning my own routes. It may take a bit longer, but at least I know I will have the route I want to take, plus I have seen the entire route, I took conscious decisions why I placed a waypoint at a certain position, so the route is already in my head. Perhaps that comes from my flying days, but it still works for me.

I do the same. It may take a bit more time, but at least I know I've scrutinized every foot of the route and, although I still monitor constantly, I feel more relaxed during the trip. Even so, I will take it off auto many times during the day and hand steer, for various reasons.

 

[Jeff F]

After having been passed within feet by boats traveling at a high speed with apparently no one at the helm, I always set a course that would take me about 5 miles from the course that most everyone else was on. I think the same phenomenon might happen with auto routing.

That's one of my general objections to auto routing. On commonly traveled routes it can crowd traffic unnecessarily into a narrow lane.

It took me a while to recognize this. Now I see people auto-routing their way around the loop following a pre-determined parade route. They're surprisingly easy to avoid once you're onto their tricks.

 

[Helmsman]

I use autoroute to get a quick time and distance check. If the route is straight forward with few hazards, I sometimes work through the waypoints deleting many and moving away from buoys and the “wrong side” of the channel. About every third route I find a problem with going over land or through shallow water.

I would never run one without checking the entire route. So I do more manual planning most of the time.