Weird Routing Behaviour (2)

[jfheath]

It's good to know that there are a few more people experiencing this and that other ways of getting round the problem are being explored. (Actually it's a bad thing that it is a widespread problem, but you know what I mean !).

I'm going to chip in a few comments at this point because I am getting a bit confused.

First of all @Peobody is absolutely correct in pointing out that the satnav is quite entitled to ask you to make U turns in some situations. It doesn't mean that there is a fault. It is often user error in plotting the points, or riding along different routes when there is clearly a 'better' way for the XT to get you to the next route point. When riding it is very difficult to know whether the XT is stuck in RUT loop from which there is no escape, or whether it is just making a sensible suggestion, given the information that it has been provided with.

------------------------------

1. Waypoint on wrong side of the road, which you mentioned.
2. Misplaced Waypoint (one meant to be on the road but positioned slightly off).
3. A Waypoint the XT did not record as "visited".

Confusion - Well I think that I can work out what is the meaning of the words written in some posts, but I am not always sure whether the author has the same intent.

Usually if we are discussing routes, we need to know whether a point on the route is a Via Point or a Shaping Point.
A Waypoint is a special sort of point that has been created and saved before it was ever included in a route. It is a different beast, it has different properties, and sometimes it can behave differently.

eg for the XT, Saved Waypoints, when they are included in a route never have their names aletered when transferred from Basecamp to the Zumo XT. Not the case with other points.

eg In the early days of Closest Entry Point, it was clear in my tests that the programmers were heading for route points rather than the closest point on a route, and it was clear that they were treating Waypoints as a different type of route point.

It's enough for me to make the distinction when I come across issues on the XT. Does it make a different if the point was first created as a Waypoint ? So in this RUT testing I have had the same route with:
. All Waypoints, all set as shaping point (except start and end).
. All Waypoints, all set as Via Points.
. All Waypoint, some of each.
. Then repeated the above with all ad-hoc point, and a mixture of Waypoints and ad-hoc points.

---------------------------------

But even in this circumstance the Zumo XT breaks it up into multiple small segments. The other thing it does is report spurious speeds in the track log. My speed will jump from 55 to 175MPH for example.

THis is to do with the fact that when the XT gets stuck in the Repeated U Turn 'RUT' loop, it stops recording the track log. I haven't noticed this excessive speed being displayed, but I have deliberately forced it when passing through the Tyne Tunnel. The stanav signal is lost, yet the satnav still displays your speed. How is this possible ?

The answer is that it isn't possible. It displays the speed that you think you are going - ie the speed that you were going before entering the tunnel, so it estimates where youa re based on that speed. Speed limit was 40 at the time I think. I entered at 30. And then when out of signal range, I sped up to 40, and continues to the end. The stanav thought that I was much further back inside the tunnel, when I hit daylight (and satellite reception) again. The satnav works out speed by calculating distance / time. So in that couple of seconds when it got my new accurate position from the satellites it was calculating my speed from where it though I was - much further back, to where I actually am. So in that couple of seconds, I had travelled quite a distance. I exited the tunnel at 127mph. Apparently.

I mention this because it seems that when the XT gets into recalculating the new route, it does not keep up with obtaining satellite information - perhaps a deliberate attempt to speed up processing ? I don't know. Perhaps an error. But it doesn't do it for normal recalculations, and it doesn't stop recording if you have the manual track recorder running. The point is, how does it know where you are to calculate the speed, if it has turned off that function. It may be doing the same thing as it did in my tunnel experiment. Catching up.

----

Basically, I created a 200 mile route around North Wales using MRA with start and finish waypoints and shaping points between. I uploaded this to Garmin drive as a GPX 1.1 file and it synced normally and calculated on the XT.
The first issue was that it wanted to navigate me really strangely to the start point (but that might have been because I’d not placed the starting waypoint accurately - possibly on the wrong side of the road). I got round this by cancelling the route and then manually selecting my first shaping point as my point of entry.

MRA v1.1 Includes the Via Points and Shaping Points, but does not keep the route itslef. The XT has to calculate it for itself - so you cannot expect the XT to come up with an identical route. You can display the route if you also transfer the track and have that on the screen at the same time. V1.2 keeps the route and the Via Points, but not the shaping points. (Basecamp also has a facility for this - to strip out the shaping points - So providing that you do not allow the XT to recalculate the route, it should follow it precisiely. But guaranteeing that the XT will never recalculate a route would be rather bold of me.

However - planning a route in Scotland is likley to produce less problems that it would in many areas. There are not that many places to get you from one location to the next - although it is more likely that you would want to tell people with whom you are sharing the route that they would need to enable that feature. A fool proof method would be to give them a track, say Go! and ask them to follow the line on the map (no navigation) - or ask them to convert it to a trip (I have never tried to share such a Track-Trip from the XT).

If you want to send me a copy of your Wales route, it would be interesting to take a look and see if I can spot any issues that would create the behaviour that you have experienced. I'll send you a PM with my email address.

Finally

The fuel pumps or the restaurant parking location is just far enough away from the coordinates of the Waypoint that the XT doesn't record the stop. @jfheath has an ingenious solution to this which I have adopted. That solution is to move the Waypoint to a spot on the road shortly before the actual location. It should be close enough that you can see the facility but on the road such that you travel through it before turning in to the facility.

I think that you have made that solution all by yourself ! Its not something that I do. Not quite, anyway.
What I described was something similar:

I put the Via Point on the road AFTER the intended stop. This is about a mile up the road that I will be on whether I detour for a coffee break or not. I usually create that Via Point as a Waypoint - I give it a name that means something to me, and save it. Then I use it when I create the route. This is because I need to be able to recognise what it is and why it is there. Making it into a Waypoint guarantees that the name is not changed during transfer.

Having the Via Point after the stop means that if I accidentally stop the route when fiddling with the XT in the cafe, I can relaod the route again. Because it is a Via Point, the XT will list it in the route points that are shown when a route is loaded and Go! is pressed - when it asks you to select the next destination.

I put a shaping point at the precise location of the coffee stop. I know that if I don't want to stop for coffee, and I carry on towards the Via Point, the satnav will complain and ask me to go back - which I ignore. As soon as I get near to or join the magenta route before the Via Point, it will stop trying to get me to go back to the coffee shop shaping point.

Sometimes I also put a Via Point just ahead of the possible detour. Create a Saved Waypoint and Label something like "choice - coffee or stay on B1234" so that when I approach the turn off for coffee it says "Arrivg at choice - coffee or stay on B1234". (This is usually too much trouble and I only use it if I do not know the area. But it is handy sometimes).

[Mzokk]

The answer is that it isn't possible. It displays the speed that you think you are going - ie the speed that you were going before entering the tunnel, so it estimates where youa re based on that speed. Speed limit was 40 at the time I think. I entered at 30. And then when out of signal range, I sped up to 40, and continues to the end. The stanav thought that I was much further back inside the tunnel, when I hit daylight (and satellite reception) again. The satnav works out speed by calculating distance / time. So in that couple of seconds when it got my new accurate position from the satellites it was calculating my speed from where it though I was - much further back, to where I actually am. So in that couple of seconds, I had travelled quite a distance. I exited the tunnel at 127mph. Apparently.

This is a slightly different behaviour. This track log was recorded with the bikes cruise set to 70mph (GPS) 74mph (bike) on a stretch of dual carrageway with clear view of the sky and no overpasses. Most peculiar. The average speed of 69.7mph for this section is correct as i had the cruise set. But I didn't have excursions to 122mph and 100mph respectively......its an Africa Twin

[MattW]

It's good to know that there are a few more people experiencing this and that other ways of getting round the problem are being explored. (Actually it's a bad thing that it is a widespread problem, but you know what I mean !).

Basically, I created a 200 mile route around North Wales using MRA with start and finish waypoints and shaping points between. I uploaded this to Garmin drive as a GPX 1.1 file and it synced normally and calculated on the XT.
The first issue was that it wanted to navigate me really strangely to the start point (but that might have been because I’d not placed the starting waypoint accurately - possibly on the wrong side of the road). I got round this by cancelling the route and then manually selecting my first shaping point as my point of entry.

MRA v1.1 Includes the Via Points and Shaping Points, but does not keep the route itslef. The XT has to calculate it for itself - so you cannot expect the XT to come up with an identical route. You can display the route if you also transfer the track and have that on the screen at the same time. V1.2 keeps the route and the Via Points, but not the shaping points. (Basecamp also has a facility for this - to strip out the shaping points - So providing that you do not allow the XT to recalculate the route, it should follow it precisiely. But guaranteeing that the XT will never recalculate a route would be rather bold of me.

However - planning a route in Scotland is likley to produce less problems that it would in many areas. There are not that many places to get you from one location to the next - although it is more likely that you would want to tell people with whom you are sharing the route that they would need to enable that feature. A fool proof method would be to give them a track, say Go! and ask them to follow the line on the map (no navigation) - or ask them to convert it to a trip (I have never tried to share such a Track-Trip from the XT).

If you want to send me a copy of your Wales route, it would be interesting to take a look and see if I can spot any issues that would create the behaviour that you have experienced. I'll send you a PM with my email address.

Thanks very much for this information (in fact thanks for the whole thread - a lot of learning for me in here).
I’ve always considered myself a pretty competent Garmin user (the only person I know who loved Basecamp!) but having only just bought my XT, it clearly does some things very differently to my previous Zumo’s and I need to learn it’s ways.

It would be great if you would look over my Wales route and let me know if you spot any user errors. Hopefully I can learn from it and fine tune my approach to creating and riding routes.
Thanks again.

[jfheath]

Private message sent. Select Private Messages from top right corner below the search box.

[jfheath]

This is a slightly different behaviour. This track log was recorded with the bikes cruise set to 70mph (GPS) 74mph (bike) on a stretch of dual carrageway with clear view of the sky and no overpasses. Most peculiar. The average speed of 69.7mph for this section is correct as i had the cruise set. But I didn't have excursions to 122mph and 100mph respectively......its an Africa Twin

It has got to be to do with the satellite reception. It works out the distance from the lat/long coordinates, and calculates the speed from the change in distance and the time. If you look at the fast speeds in your log, in 2 seconds you traveled 205 ft and in the next 2 seconds it was 359. Satnavs do have a tolerance and can usually be between 2 and 5 metres out of position. Sometimes it is much more. You can display the accuracy on one of the 4 trip data displays on the right hand side of the screen (landscape view - or across the bottom, portrati view). Press one of them (the white one, bottom right, is always available), and choose the data item that you want to display. GPS accuracy is down at the bottom of the list on mine and is currently displaying an accuracy of 10 yards. But I'm in my study.

But take my current error of 10 yards. That is 30 ft. That is a possible 60ft discrepancy between two consecutive readings - if one reading was out by 30ft in one direction, the next was out by 30ft in the other.
They can be pretty good at calculating the error, but its more difficult to pinpoint your precise location !

It can vary as trucks go past. There must be other circumstances too. I thought that it calculated a 'moving average' to get the speed - eg It seems to record a fix every second or so. But it will take the fix from 5 seconds ago up to the latest fix. and working out the speed from that. I don't know this. I have assumed it to be the case because when you accelerate hard or brake hard , it takes about 5 seconds for the satnav speed to catch up with your actual speed.

Have I seen some comment about the proximity of the modern TFT screen displays interfering with satnav behaviour ? Certainly the heated windscreens that some Ford cars have can stop it from receiving a decent signal.

[Mzokk]

I think you might be onto something. I looked up the old Logs on my Nav4 when it was on my 2014 BMWr1200RT and there is no odd behaviour with jumping speeds. Same thing with my Zumo 340 and 390 when attached to my old R1150GS However, the Nav4 logs When it was used on my 2020 Honda Africa Twin Adventure Sports exibit the jumping speed thing in a similar way to the XT. The Irony is that the Honda TFT on the Africa Twin is made by......GARMIN!. The screen has a GPS antenna built into, which it uses for time and probably to ensure good GPS signal when you have a phone attached for Android Auto/Apple Carplay. Would be a bit of a kicker if two Garmin devices interferred with each other.

Likely the reason for the multiple track logs too!.

[Stu]

Today I noticed that mine was trying to take me to the last turn I made and not the last turn we missed

We pulled out of a junction then 100 yards up the road we missed the turn! we turned around and the sat nav was trying to take us back to the last junction we pulled out of to then start the route again

[jfheath]

That behaviour is typical if it has already triggered the repeated U turn scenario. You cannot tell whether or not it has been triggered until after the first time that it happens. And you can usually confirm it the second time it happens. It is easier to confirm by the magenta route if you have disabled U turns.

It is usually after a previous deviation after a skip or strating with Closest Entry. But I think I have seen it at other times to. Each time it recognises that you are off the route, it 'marks' that point. Any future recalculation only calculate the route up to that point. I think it is a time saving device. If you ignore it, it will do the same again, but only calculate back to the point where it last told you to turn.

You can tell if it has been stuck in a RUT - the automatically recorded track log (CurrentTrackLog.gpx) will be broken at every point where it has recalculated.

[Stu]

I thought I would try some testing today!

The plan was to do a route on my route app 4 times to see if I got any issues when I missed a turn

I did the same route for all 4 the first one loaded and driven as it should be missing a turn without editing/touching the route

The second to recalculate the route and see if it was different

The third to edit the route on the zumo and see if it made a difference

The forth to drive it in car mode to see if there was a difference

All routes behaved as it should re routed me a different way to get me back on track so a total waste of time

I just don't think there was enough options for it to think about to be honest so more testing is needed

[jfheath]

That is as it should be. Not a wasted effort - it actually confirms something that I hadn't tested properly.

At one point I got so engrossed with trying to find out what made the route develope into a Repeated U Turn loop (RUT), that I nvere bothered testing how to make it work. The test engineers had told me that they could not get thier units to behave like mine. So I wanted to be able to guarantee that the XT would alway get into a RUT, to pass on that information.

So it came as quite a surprise, when at one point I cancelled my test route, and started again - building the route on the XT using the same save Waypoints. The route behaved perfectly - recalculating correctly and earlier than I would have expected.

To make it go wrong I had to have the same route with an earlier point that I did not wish to visit. And press skip.
Or have a route and start it part way round by selecting closest entry point.

After that, if I deviated from a route and didn't obey the instruction to go back, it would no longer find the best way to get tot he next route point. It would focus on getting me to go back to the point where I had deviated. Each time I ignored an instruction, that became the next point to go back to.

But it seems that if you don't skip a route point (by tapping skip), and it you don't select closest entry point - then the XT behaves 'properly'.

Which begs the question - what do you do if you have to skip a route point to get past a road closure ? I have workarounds - which largely involve following the track instead - or possibly reloading the original route and removing the points up to your next route point.