Weird Routing Behaviour (2)

Having Garmin zumo XT problems? there is loads of help and advice in this forum
jfheath
Posts: 2276
Joined: Sat Oct 19, 2019 4:17 pm
Location: West Yorkshire, Uk
Has liked: 281 times
Been liked: 613 times
Great Britain

Re: Weird Routing Behaviour (2)

Post by jfheath »

Oop North John wrote: Sun Apr 23, 2023 12:03 pm My observation of the 3 u-turns then routes sensibly is based on a route that previously it would take me within a couple of miles of the destination, still wanting me to u-turn, and do about 8 miles back along the same roads I'd come along.
My examples went on with U turn request until I was within less than a mile of rejoining.
Is your track history playing a part in this. I noted the other day, that the XT decided to calculate the route down my preferred route, rather than the on that used faster roads as it normally does.

In your example did you press skip / choose closest entry. I find that it requires one of these some time (maybe a long time ) before deviating from the route.

Today, I discovered something really important. I found another way that doesn't work.
jfheath
Posts: 2276
Joined: Sat Oct 19, 2019 4:17 pm
Location: West Yorkshire, Uk
Has liked: 281 times
Been liked: 613 times
Great Britain

Re: Weird Routing Behaviour (2)

Post by jfheath »

Today I did an impromptu test.

I put in a route with a start, and point soon after the start that I could skip and a destination in Ilkley. Only about 8 miles, and I was passenger in the car so I could watch it. The route was to Harrogate. I had no intention of going to Harrogate.

I went throught the start and then skipped the 2nd route point that I created specifically to skip - just after the start. The route recaclulated to Harrogate.

At the roundabout the route went straight on, I turned right. The route recalculated ahead to find the other obvious route to take me to the A59 towards Harrogate.

I ignored the instruction again. It recalculated. I expected it to take me back to the roundabout, but I had U turns disabled. It didn't.

It routed me ahead, and then found a side road to take me back on a parallel road to the turn off that I had just missed. THis was what I wanted - A visible loop back. With a RUT, that loop back will remain in place
I ignored this, and the satnav routed me ahead to find another side road it could use to double me back. The loop back that I expected to remain in place vanished. It had calculate a brand new route to take me back to join the road to take me to the A59.

I ignored that instruction. And then, massive surprise when I ignored that - it routed me ahead along a completely different road to take me towards HArrogate.

At no point did this behaviour differ from what I would have expected my 590 or 595 to have done.

There was no break in the recorded track log - typical of the behaviour when a RUT is caclualting.

I mentioned earlier that when I thought I had update Express, I came back and found my Zumo needed restarting. It needed restarting 3 times, at one point after saying it was installing software of the startup spalsh screen.

The unit still says v6.50 but there are other signs - like apps being put back to original positions that indicate that something serious was happening.

I need to reset this unit back to factory. Scrub all of my track logs and riding history and take this out on one of the test runs that until now had always failed. The one of the pictures.

But this behaviour is brand new. I was certain that the XT would get stuck in a RUT.

I'll do this test as soon as I can - but if this is now fixed, I'm really pleased.

---

I just went to take copies of my track logs. More odd behaviour. aAst time I took copies up to 6.gpx and CurrentTrackLog.gpx

Now I have a 6, 7 and 8. - which I would expect - but 6.gpx has grown from 438Kb to . 1,100Kb CurrentTrackLog.gpx at had grown to 2.75Mb

It may be that CurrentTrackLog keeps growing the file until it knows that it has finsihed with a day's trip, and then dumps it into the archive - adding data to files that have already been saved there.

Today, I discovered something really important. I found another way that doesn't work.
danham
Posts: 273
Joined: Fri Jul 13, 2018 2:25 pm
Location: Cape Cod, MA
Has liked: 87 times
Been liked: 83 times
United States of America

Re: Weird Routing Behaviour (2)

Post by danham »

jfheath:

I have been doing my best to follow the twists and turns of the RUT saga and appreciate your thorough reporting. With the disclaimer that I may not have enough coffee in me yet this morning, may I pose a question about how best to avoid RUT when navigating a complex, pre-planned route?

Here's how I minimized this kind of error on my zumo 660 and I'm hoping it or something similar will work for my XT when it takes its first big (3000-mile) ride in June. It boils down to allowing U-Turns in Settings so that the GPS does not offer confusing go-arounds to mimic a U-Turn, ignoring the command to do the U-Turn, making sure it is set to ask before recalculating (or Never), and using the purple stripe to manually navigate back onto the intended route after a wrong turn. The ability to display the track on the XT should make this even easier, right? It seems like that and not using the Skip function should do the trick and allow the XT to pick up where it left off, without wrecking the route, once it realizes it is back on the purple stripe. Yes? No?

A typical 200-mile route for my big ride has maybe 20 Vias (stops) and an equal number of shaping points.

Am I missing an important point? Suggestions welcome.

-dan
Zumo XT, 660, nuvi 760 and many retired units dating back to the GPS III+
2018 Kawasaki Ninja H2 SX SE
FrankB
Posts: 414
Joined: Mon Aug 10, 2020 7:22 am
Has liked: 91 times
Been liked: 125 times
Netherlands

Re: Weird Routing Behaviour (2)

Post by FrankB »

@danham
Yes, thats how I do it. If it stll fails, restart the trip from 'closest entry point'
User avatar
Peobody
Subscriber
Posts: 1333
Joined: Tue Apr 20, 2021 1:33 pm
Location: North Carolina USA
Has liked: 99 times
Been liked: 277 times
United States of America

Re: Weird Routing Behaviour (2)

Post by Peobody »

@danham ,
The XT won't navigate you forward when you bypass a via point (alerting waypoint) without first skipping it. Whether this could result in a RUT scenario later in the ride has yet to be determined (AFAIK). A more important reason for trying to never skip a point is that the remainder of the route gets recalculated. I don't think I have seen a recalc not mess up the planned route. Having its track on the map is invaluable in identifying that condition. Depending on the circumstance, you may be fine continuing on by following the track. Otherwise, a restart of the route is in order.

I keep u-turns enabled for the same reason as you. I have recalc set to ask first. Unfortunately, a skip doesn't generate a recalc prompt.

I took a ride yesterday using a route that was created with a lunch stop. Before leaving, I removed the lunch stop point by editing the route using the XT. I didn't realize how much that messed up the route until I was well into it. Thankfully I had the track of the original route on the map.
2008 Honda GL1800 Goldwing
1995 Kawasaki ZG1000 Concours
zūmo XT linked to Cardo Packtalk Bold and iPhone SE.
Oop North John
Posts: 296
Joined: Fri Jul 27, 2018 10:27 am
Has liked: 77 times
Been liked: 44 times
Great Britain

Re: Weird Routing Behaviour (2)

Post by Oop North John »

jfheath wrote: Sun Apr 23, 2023 10:09 pm
Oop North John wrote: Sun Apr 23, 2023 12:03 pm My observation of the 3 u-turns then routes sensibly is based on a route that previously it would take me within a couple of miles of the destination, still wanting me to u-turn, and do about 8 miles back along the same roads I'd come along.
My examples went on with U turn request until I was within less than a mile of rejoining.
Is your track history playing a part in this. I noted the other day, that the XT decided to calculate the route down my preferred route, rather than the on that used faster roads as it normally does.

In your example did you press skip / choose closest entry. I find that it requires one of these some time (maybe a long time ) before deviating from the route.
I chose the first (only) via / way point. Interestingly the track logs stop / start at each recalculate point, but no massive speeds / altitude are shown, less of a gap between the track legs as well, compared to a month ago..
jfheath
Posts: 2276
Joined: Sat Oct 19, 2019 4:17 pm
Location: West Yorkshire, Uk
Has liked: 281 times
Been liked: 613 times
Great Britain

Re: Weird Routing Behaviour (2)

Post by jfheath »

danham wrote: Mon Apr 24, 2023 11:38 am jfheath:

I have been doing my best to follow the twists and turns of the RUT saga and appreciate your thorough reporting. With the disclaimer that I may not have enough coffee in me yet this morning, may I pose a question about how best to avoid RUT when navigating a complex, pre-planned route?

Here's how I minimized this kind of error on my zumo 660 and I'm hoping it or something similar will work for my XT when it takes its first big (3000-mile) ride in June. It boils down to allowing U-Turns in Settings so that the GPS does not offer confusing go-arounds to mimic a U-Turn, ignoring the command to do the U-Turn, making sure it is set to ask before recalculating (or Never), and using the purple stripe to manually navigate back onto the intended route after a wrong turn. The ability to display the track on the XT should make this even easier, right? It seems like that and not using the Skip function should do the trick and allow the XT to pick up where it left off, without wrecking the route, once it realizes it is back on the purple stripe. Yes? No?

A typical 200-mile route for my big ride has maybe 20 Vias (stops) and an equal number of shaping points.

Am I missing an important point? Suggestions welcome.

-dan
That is a pretty good way to do it. I'd reduce the number of Vias and change some to Shaping points - but don't do it on the XT - it will likely move and rename them ! Oh yes !

The reason for having more Shaping points - I suggest at stopping places that you will definitely visit - is that if you wander off route and rejoin it and all you have missed out is Shaping Points, it will continue to navigate you ahead. If you miss out Vias, it will take you back to them, even if you have rejoined the route.

Closest Entry Point also seems to have an effect on the RUT behaviour (Ie it might trigger it in the same way that Skip does) I have examples of this happening, although I would stop short of calling it evidence yet.

One thing that you can do is (if you know wher eyou are in relationt o the shaping points) choose the next shaping point that you want to aim for, and change it to a Via Point on the XT. (That switch works OK) Then start the route and select the new Via as the next destination.

I've just come back from a four seasons tour of about 2 hours. 4 Seasons because it was nice and sunny when I set off and on the way it rained, hailed, sunshined and snowed. Got down to 2degC at one point.

And the sanav was showing me some new tricks that it had up its sleeve.

More of which in the next post.

Today, I discovered something really important. I found another way that doesn't work.
jfheath
Posts: 2276
Joined: Sat Oct 19, 2019 4:17 pm
Location: West Yorkshire, Uk
Has liked: 281 times
Been liked: 613 times
Great Britain

Re: Weird Routing Behaviour (2)

Post by jfheath »

This is an issue that I came across today while testing out a tried and tested route for observing the behaviour of the Repeated U Turns.

To understand this, you need a picture. And you will need to click on it to get the full size image.
Route Map.jpg
Route Map.jpg (113.32 KiB) Viewed 652 times

The picture shows the BAsecamp route that I loaded into the Zumo. The yellow circles are all Via Points.
Nos 1 & 2 , bottom right corner are the start point and a 2nd point close after that I intend to Skip after passing through the start point.

Point 3 is a token gesture that I am taking the route north along the B6160.
Point 4 is a Via Point on the road I intend to take, leading away from Hawes - The B6255.

Note Ingleton - is a town at the SW end of the B6255.

Point 5 is at a hamlet called Cleatop Just after the point where My Route and the A65 join up.

The dotted red line is the route that the XT calculated after I pressed to SKIP point 2. THe entire route was recalculated, The new route visited point 3 on the B6160 and then tunred round to go along the A65 all the way to Ingelton, and then turn NE to pick up the Via Point #4. I have put the dotted red line below tha A65 - Point 5 is not part of that section of the route.

The XT's route will turn round just after Point 4 and follow the same route back.

When navigating the route - I got stuck in a RUT loop shortly after passing Point 3. However, that is not unreasonable. It probably IS quicker to turn round and travel the A65 to get to Point 4. That continues to be the case according to the XT routing (setting the bike position in simulation mode and seeing which route it takes to get to the Via Point.) From Conistone, the route north becomes the faster way.

However, it has already entered a RUT loop. I had U turns dis-allowed, so it had to use side roads to route me back, and these could be seen in the recalculated route - not taking the main road back, but deviating into all of the side roads where it had attempted to redirect me.

Beyond Kettlewell, the route was plotted ahead. And I do not understand this. The milage to the Via was about correct, but the XT had plotted a route NE from Buckden, and I was heading on my original route NW through Yockenthwaite. Immediately the XT tried to loop me back on the white roads E of Hubberholme, and abck to the A65 - or back to the point of deviation. I don't know which.

It then all became academic. The distance to Via Point 4 I reckoned was 10 miles. The Trip data on the screen showed 42. 32 miles more. Yet it had now decided to route me ahead again.

Now that was odd. It is about 15 miles from the via Point 4 to Ingleton.
Hubberholme.jpg
Hubberholme.jpg (39.17 KiB) Viewed 652 times

In Hawes the satnav was leading the way heading me towards the Via Point. Here is is on the screen, the Via Point about a mile ahead. (Don't ask me how the Explore green flags got on there. I didn't set it up to use Explore). But look at the distance to the Via Point. And it is the Hawes Via Point the XT is heading for - I checked by tapping skip and then saying No.

Via Pt 4 Mile ahead.jpg
Via Pt 4 Mile ahead.jpg (45.09 KiB) Viewed 652 times

There were no announcements - approaching / arriving at - neither spoken or displayed. This is the screen just after I went through the Via Point.
Through Via Point.jpg
Through Via Point.jpg (41.5 KiB) Viewed 652 times
Notice the distance to the Via Point and the distance to the turn. I hadn't got a clue where Burnmoor Crescent is, but I guessd Ingleton. Is the stanav heading for its route at Ingletone ?? That didn't make sense.
No the road is just wide enough for two cars to pass. Dry limestone walls lined both side and there is no place that the map showed where it could turn me round to head back. So 15 miles to Burnmoor Crescent. (Burnmore Rubber ?) and 15 miles back.

Hmm. Is Burnmoor Crescent in Ingleton ? Yes it is. It is the first place that the XT had the opportunity to turn me round using roads plotted on the map.

So I pondered on this for a while. Why ?

My route was planned not to go through Inlgeton at all. My route was planned to take an anti-clockwise route. The XT planned a different route - there and back. And it planned to approach route point 4 from the oppositie direction.

And I reckon that is why it was taking me 30 mile out of my way to visit a point that I had just passed through. I was on the magenta line - it should have recognised that. I passed throught he via point - it should have recognised that.

But it didn't - because I was heading int he wrong direction - and that makes sense.

But this is something to be aware of - it has nothing to do with the RUT. It is routing according to its recalculated route that came about from pressing skip. And it recalculated the entire route. And in itse route, it had to get to point 4 from the SW, not from the northeast.

And that means that it has to be taken into account when route planning. If you do like I did and have not route points between 3 and 4, then the satnav is free to choose its own route, and that may result in the XT expecting to hit the point from a different direction from what you expect. And that will certainly affect which roads it takes.

But I established before any of this, that the XT is still getting into a RUT - and it did it as soon as I left point 3 heading north.
And this can be seen in the route back to the A65 from Cracoe. Going back is the quicker way from there, but it is still following the detours that I ignored on the road towards Cracoe.
There were 3 successive detours suggested in this area. All ignored. The main road back is the yellow line. THe kinks are taking me back through the detours that it suggested.

Cracoe RUTs x3.png
Cracoe RUTs x3.png (155.92 KiB) Viewed 652 times

Today, I discovered something really important. I found another way that doesn't work.
MattW
Posts: 3
Joined: Tue Apr 04, 2023 5:45 pm
Has liked: 7 times
Great Britain

Re: Weird Routing Behaviour (2)

Post by MattW »

Very interesting 👍
(And that very loop around the Dales is one of my favourite 'day out' rides :) )
User avatar
Peobody
Subscriber
Posts: 1333
Joined: Tue Apr 20, 2021 1:33 pm
Location: North Carolina USA
Has liked: 99 times
Been liked: 277 times
United States of America

Re: Weird Routing Behaviour (2)

Post by Peobody »

I am not surprised by anything that you experienced. The RUT issue remains a problem. Very common for me is the recalc that results in a diversion to a waypoint that was placed for route control purposes but was close enough to an intersection for the recalc to divert you to it and then return back to a new route rather than riding through that waypoint as originally intended. As you mentioned, another point between 3 and 4 might have resulted in a more reliable recalc but we should not have to try to out-think the XT when planning routes.
2008 Honda GL1800 Goldwing
1995 Kawasaki ZG1000 Concours
zūmo XT linked to Cardo Packtalk Bold and iPhone SE.
Post Reply