Zumo User Forums

Thanks that was helpful.

Surely to a Yorkshireman it should be "Rooooooooooooooooooooooooooot"! Joe is the man

Shaping points seem to like faster roads, and can change how they want to

Using GPX files with the Zumo XT2 - Possible Solution ?

I have mentioned this topic before. It seems that if you put a gpx file of a route onto the Zumo XT2 - say the SD card - and the route points are accurately placed, then when the route is imported the XT2 will retain the position of the route points - via and shaping - and the route will pass through every route point. It may be recalculated, but the route honours the route points. Providing that the Tread App is not allowed to see it. (eg Tread is not loaded, or BT to the phone is turned off).

But there is another way that is worth investigatingwhich so far I have yet to provoke a failure. Not one that I can't explain away or one that I can duplicate, anyway.

So to get a GPS file from any source (including Basecamp) to work :-

1. Create the route and save it as a GPX file - as normal. I still use Basecamp.
2. Send the gpx file to your mobile phone that contains the Tread app. I use email from my PC to my mobile phone.
3. Open the GPX file on your phone with the Tread App.
4. Let it synch with the XT2.

The route will probably be recalculated, but it passes through every via point and every shaping point.
And by default it has the flag set that prevents the route from getting into a RUT scenario.

At the bottom is a video of this in action. It's about 16 minutes with 3 chapters.

If you can't bear to watch all of it. Then links to each of the three chapters are provided below. You may want to see just the last chapter.

00:00 Opening Page
00:13 Ch 1 - A description of the investigation, the screen and a test route.
03:05Ch 2 - Using a GPX from the SD Card
07:36Ch 3 - Sending a Route by email to the Tread App on your phone.

This method is well worth a try. It just may be the GPX solution that we have been looking for.

I have followed the above method, making a shopping trip to Harlow and return routes with tracks.
On my XT2 the outward journey follows my intended route with 3 detours with obvious faster routes.
The return trip moved shaping points as seen by the dark grey tracks position.

Calculation mode: Faster time
Off Route: Automatic
Is the XT2 set right?

Also, of note.. each gpx file was imported individually and on the second import, although I selected the first Collections Folder, I ended up with 2 Collection folders of the same name.

Thanks for those files.

Interesting. A few observations - which might contribute to the moved shaping point. Or not.

The track that you included does not follow the roads exactly on my 2025.2 map. So I had to recalculate the routes.
There were a few differences from your track in Basecamp and my route as a result.

We know that all route points move slightly. Garmin seem to be catering more for the inaccuracy of people trying to produce accurate route maps on tiny screens with small fingers. So it seems that most points in a route are subject to slight modification. When placing a route point on the tread app on the iPad and Android phone - you place a point - it pops up a list of things that are nearby and you get to choose one of them. This places the point accurately on a road. It also tells you whether or not the location is supports for motorcycle routes. Routs containing route points seem to undergo a similar validation check ! But it seems to move them to places that are on the route that you planned.

However - what you point out reveals some really interesting behaviour when it cannot actually follow the route that is plotted. I'm talking about near the roundabout on the B1007 at a junction with Longstomps avenue - SW of Great Baddow - near the easter end of your route home.


I have drawn a red and a green line. Just above, circled, it goes down Dove Lane. There in no road that links the B1007 and Dove Lane. There is a cycle and pedestrian path. But no way that the satnav can calculate a route for the motorcycle. So it either has to give up, or esle come up with an alternative.

Your route lines and my maps did not coincide - so I recalculated the route before sending it to the XT2. BC produced a route that followed the red line - doubling back and going through the fixed route point on Linnet Drive.
The XT2 did something different. It decided it couldn't get to that shaping point and moved the shaping point to where the black circle is - and plotted the green route.

Whether or not this was due to the map that your plot follows not being quite on the mapped roads - I don't know. But we do know that under certain circumstances, when the XT1 recalculates a route, its behaviour changes. It seems to be a different kind of route - where the route takes priority over shaping points for example. It will let you completely ignore Via Points - and they disappear as you drive past. But then it changed on my route too - and that didn't try to take the cycle path.

So you are right. It moved the route and the shaping point. But it raises some questions for me to ponder.

So just thinking out loud here. I'm not asking these questions of you !!

Did it know about the original route. If so did it try to keep close to it ?
If it only had the shaping points to go on, why did it move one of them ? Would it have done it if there were two shaping points.
Would it have done it if there wasn't a route error.
If it did know about the original route and that it was impossible - did it then recalculate it using a different set of rules.
I recalculated the route before sending it to Tread - so as far as the route was concerned - the Tread app didn't receive any incorrect routing.

I don't mind it changing a route between shaping points. Your routes did that in a few places. That's down to the person doing the route planning to place points more appropriately. (Not you or me in particular - just in general.)



Thanks for that input - food for thought, and very useful.

One thing to remember when using Basecamp to generate the route, it uses the good fastest routing algorithm where as Tread and the XT2 use the faster roads, so it is likely that more shaping points are needed .

Edit - when I upload a route as per JFHeaths method I review it, and possibly change some via points to shaping ones. But, sometimes the updated route misses the road I was on, and generates a very new one. If I then undo the change, and then redo it, the shaping point and route stay on the original road, SOMETIMES! I think this is the behavior I see in my Tread Quirks thread.

Big route change in the attached file which doesn't stay on the original if you change "Rey Cross" to a shaping point in Tread.

Edit #2 - I let the XT2 sync with Tread so that the below route was in the XT2, all shaping points and then changed Rey Cross to a shaping point in the XT2, and it moves, but the route is the same as before when it was a via.

So it might be that any changing from a via to a shaping point should be carried out in the XT2?

Thanks John

I am trying to leave routing changes due to different algorithms out of the equation for now. As you say - they are fixed by more prudent use of Shaping Points. I'm focusing on trying to make sure that whatever route it produces, the Sh Pts stay in the same position - or in a position that doesn't affect the route.

I've learned not to change Via to Shaping and vice versa - for two reasons
i - The route point moves - usually backwards, often onto a road that joins the two adjactent point by the fastest route
ii - The XT started treating them differently. Put a shaping point in, it would go through it. Change it to a Via it would often double back to find a faster road.

Changing the Rey Cross point in the Tread App on the phone: The point moves but there is only slight movement of location. No route change. I haven't tried changing it on the XT2.

I've yet to play with the second scenario you describe - but that sounds familiar. I'll check it out.

I'm not clear about what is going on between XT2 and Tread App route.

I have an 'inkling' forming. A twinkling in the eye of an idea - but it hasn't got round to developing into a fully formed guess:

To describe this inkling, jfheath wrote:
If the route comes from the Tread app - then the Zumo treats it as not imported. Its a bone fide route. The two - Tread and XT2 - are designed (it seems) to work that way.

If the route originates from the XT2, then it could have come from anywhere. It is treated as imported and is susceptible to RUT behaviour. Also, route points could have been placed anywhere on a completely different map. Such routes need to be fettled by the soft logic software which puts them on a nearby road.

If you change a route on either device it then has to synch to the other device.

Q. If a route that originated in the Tread App is subsequently edited by the XT2, and it synchs back - does tread treat it as a rogue route that needs to have soft logic applied to produce a route that is plotted on roads, rather than through someone's front garden.

Q. Does plotting a route on a parallel road eg A6 / M6 invite that logic to be applied. Maybe it thinks....It's on the way between Via Points. It's only like moving a point from someone's front garden onto the road.

I have seen some odd behaviour when I have been making changes (I change the name of the route by putting a full stop or two at the end). That forces synch between tread and zumo - and back, it seems. I like to wait for a complete pair of synch symbol to show: ie the orange Explore icon with an arrow poking it from underneath, followed by a blue rotating circle. I don't know what each means but it is after seeing that pair, that the route changes. I think the orange icon means it is sending data to the Explore database for Tread to pick up. I think the blue circle is the XT2 loading data from the database. But I don't know.

Anyone that knows anything about what is really going on could probably pull those ideas apart. But as I say - I'm just gathering information - sitting back and watching what happens and when. Slowly they get assembled into something useful."

@jfheath

Hi John,

Just an idea I had, cant test it myself, because I dont have an XT2, but maybe it's worth testing.

I had a look at this video, where they pair the Tread app with a Zumo. (I think its an XT2, but not sure)
At the end I see the same question as with Garmin Drive App and XT1. Enable Garmin Explore? Have you ever tried 'Maybe Later'?
I would hope that you can transfer a GPX wireslessly and 'enjoy weather and traffic', but Routes will not get recalculated etc.



Frank

For some reason my maps were not available and the route was quickly planned with UKI and yes, more route points certainly would have helped.
On the XT2 the route there was OK until the 5th route point (The Plain) where it wanted me to double back and take the Woodside shortcut road to Thornwood instead of carrying on to the junction and turning right onto the B1393.
On the return journey at the 5th roundabout where the route should have turned right, that is a hamburger style roundabout which is clearly signposted way back down the road, No Right Turn! Trying to report a map error just produces a white blank box.

I'll try another route to test.

jfheath wrote: ↑Mon Nov 04, 2024 3:38 pm Thanks for those files.

The track that you included does not follow the roads exactly on my 2025.2 map. So I had to recalculate the routes.
There were a few differences from your track in Basecamp and my route as a result.

Thanks for that input - food for thought, and very useful.

More route points? I hate having to 'lace up' a route with loads of shaping points. It shouldn't be necessary, and using the points that you had clearly showed that mostly they were staying in place - only the route was changing. And that's ok. The place where the original route seemed to take a cycle path was interesting. Because moving that shaping point was actually a pretty good solution to that problem. Whether it knew that the route was plotted on the cycle path, I cannot tell. But it got me thinking.

The XT1 does something quite clever (maybe the XT2 does this too, but I haven't checked). Once you have indicated that you are not following the plotted route exactly - eg by pressing skip - It treats Via points differently. It still navigates to them, but if you deliberately ignore instruction to take you towards them, it simply removes them from the route. The flag disappears, and it continues to the next route point. I posted a video on here of it doing just that.


It can be difficult to separate

* odd routing behaviour - eg like the XT1's preference for main roads
* normal route variation - ie different results but passing through all of the Shaping points
* Slightly moved shaping points that do not affect the route - which ensure that points are on a valid road

and

* Routes which change and take the shaping points with them

The first three are down to the user planning the route. The last one is the one I am trying to learn about.

I try to find places wher alternative routes are limited. Hence the A6/M6 example in the video.
My RUT testing was done mainly in a triangle where one of the Yorkshire 3 Peaks (Ingleborough) forced the satnav to take just one of two alternatives.

Another revealing one is to plot a shaping point in a village that had been by-passed.


The satnav will always want to take the motorway, the bypass, the faster route. So a single shaping point half way, or two shaping points at 1/3, 2/3 distance will force the issue one way or another with nothing silly getting in the way. I've not tested any of these on the road, like I had to with the RUT behaviour. Its not necessary.

I've updated my Android Tread to 2.09.51 this morning and the only consistent way Tread handles changing of via points to shaping points is it is inconsistent!

It still effectively removes "Rey Cross" sometimes as a point in the route, and heads down the M6 instead of still along the A66, quite a major change of routing.

But if I undo the change, and the redo the via to shaping, and repeat a few times the Tread generated shaping point seems to move closer to where the original via point was.

My head hurts!