Zumo XT Glitches

[jfheath]

I wonder if anyone has observed or can explain the following glitches that I have observed in the Zumo XT behaviour.

I have updated the Software to 2.50 (current 5 May 2020), and have also updated the subsequent small update (also designated 2.50 which includes a fix for some Closest Entry Point issues). The maps on The Zumo are up to date, and match the maps that I use in Basecamp.

I am no novice with Zumos - and for a while had well regarded document available on this site explaining the stuff that couldn't be found in the manuals regarding the 590 and Basecamp. Although no longer posted, that is still available for anyone that wants it. Just PM me.

viewtopic.php?p=3205#p3205

----------

So on to the XT issues.

1 The moving Shaping Point.
On the Zumo XT Trip Planner App. Build a very short route with a start point, an end point. Make these so that at least a mile away from the start and the end there are at least two possible routes that can be taken. Eg a main road / by pass and the old road through a village, or over a hill. So both routes will start along the same road, and finish along the same section of road. Then insert another route point in the middle of the slower route to force it over the hill, or through the village. Save it and look at the map of the route.
Now on the XT, change the middle point to a shaping point - you just press on the flag and it changes - there's a warning message, and look at the map again. Every time I have tried this, the shaping point is no longer in the place that I placed the Via Point. It has moved and changed its name. So far, it has always moved onto the road that is the faster route. It forces the navigation to be along the route that I didn't want to take.

2 The Closest Entry Point glitch

This was much improved with the patch that came after 2.50 (They may have included it in an updated 2.50 software launch, but when I upgraded I found the patch was available a couple of days later. Previously it seemed to find the closest Via Point or the closest non-alerting Waypoint. But it seemed to completely ignore Shaping Points. (Up until then, I had thought that a non alerting Waypoint / Favourite was simple regarded as a Shaping point for routing purposes). !

However, the fix sorted all of that out and now it seems to find the place on the magenta line which is closest to the current position of the bike, navigate to there and then continue with the original route from that point. It seems to work well.

Until - I positioned the bike about a quarter of a mile from the original route and Closest Entry Point ignored that and went heading off down the road, missing out important routing points and re-joining the original route 100 miles or so later. The only thing I can think of is that part of the route involved a vehicle ferry. (I had no avoidances set). There was a Via Point at the ferry. There were shaping points, but it took a 22mile longer route to get around the estuary. I don't understand that given that it had to go through the nearest point of the original route a few hundred yards up the road.

3 Changing a later section of the route after Skipping an early route point.

In my experience, this doesn't happen. I tried the same route on my 590 and it didn't do the same thing. I had a test route with 4 Via Points, one of which #3 was well out of the way - it would have involved a 5 mile longer journey than the one I intended to take. The #3 Via Point was in the towards the end of a 5 mile detour. Another Via Point #4 was placed just after the point where the detour met the road I intended to drive. Those two Via Points #3 and #4 were about a mile apart. Early on in the route, I skipped the very first Via Point #1 before I reached it. There was another Via Point #2 after that one The route changed between #3 and #4 so that instead of plotting the 1 mile between the two, it doubled back, retraced the track to where I had ignored the instruction and approached the #4 Via Point from the direction we actually drove. It chose to turn round, do 4 miles instead of taking the shortest, quickest 1 mile road between #3 and #4.

I put in a shorter route taking in the same two points, to see how the satnav would calculate it. It took the shortest, fastest route. Something happened as a result of pressing the Skip button. I repeated the same thing at my desk using the simulator.

But, as I said, it doesn't do it on my 590 which uses the same maps.

Any suggestions about any of the above, or similar experiences gratefully received. I'll phone up Garmin see if I can find a way to send the evidence.

[jfheath]

I finally found out how to get in touch with Garmin.

After much searching on their site, they have a phone number. You get to speak to someone and they log all of the details, and then if you have documentation, they give you an email address which you have to reply to. That is important as the subject line contains special reference numbers and characters. If you use that email address without the appropriate, logged reference, it bounces back.

So I have done all of that, logged all 3 issues with them and provided screen shots and GPX files.

Would you believe - the Closest Entry Point fault would not repeat what it did on Sunday. I couldn't get it to go wrong, but I did have a screen shot of that, so they now know that whatever the fault is, it doesn't happen all of the time. I'll have to try it next Sunday at the same time - see if it is an issue with the Ferry that was on the route.

[CRracer]

Had you restarted your zumo, after updating, but before your issue occured, or after your issue occured?

[jfheath]

Thanks for the comment / question.

Yes, I did a reset and restart after the updates.

I am really puzzled as to why test 2 worked that final time, but not on the numerous times previously. I have contacted Garmin, and they replied yesterday. - wanting me to do the same thing. I followed their instructions, but got the same results. I will test this out again on Sunday night, which is when I did the last test. Its the only thing I can think of that has changed. Unless there has been a second patch which was installed over wifi without me noticing.

I have replied back. But i added some extra information.

Before these tests, I had carried out some tests on the behaviour of the Closest Entry Point. I had created a route which went from A to Z. The direct route from A to Z is a main road, but my route zig zagged its way to route points which alternated from one side to the other, repeatedly crossing the main road to get to the next route point.

I had the notion in my head that 'Closest Entry' was actually finding the closest route point. And I wanted to find out which route points it would pick as a start. Any route point, or only certain types of route point. (Since when starting a route, you can only select Via Points, not Shaping Points). So into my route I put Via Points and Shaping Points (both of which were created with the drag and drop Insert tool on Basecamp and made to be alerting or non alerting respectively), and I created some Waypoints with the Waypoint Flag tool in Basecamp, added those to the route, and made some of them alerting and some non alerting. So they looked like Via Points and Shaping Points (and should be treated as such by the routing software).

What I discovered was that the Closest Entry feature seemed to be selecting the closest routing point, rather than the part if the route that was closest to the bike. But is would only select my Via Points. It would not select the Shaping Points that I had plotted. Until I got to testing the Waypoints. It would select any of those, whether they were set as alerting or non alerting.

Incidentally - this behaviour seems to have been improved with a 'patch' that came out a couple of days after I installed v 2.50 software.

So it occurred to me that perhaps someone doing the coding had not applied the same definitions to the route points that had been used in the 59x series of Zumos.

So I went to my test 3 - the one which seemed to recalculate the route between two Via Points up the road when Skip was pressed at the start. In my mind it should only calculate the route up to the next Via or Shaping Point. Thats what the 590 does. I changed my Via Points to Waypoints - ie I saved Waypoints for the same locations as my Via Points, added them to the route and deleted the Via Points that had been there, and transferred the route to the Zumo. On the Zumo map the route looked identical, with flags marking the spots. But it behaved differently. This time on the simulation, when I skipped the first Via Point, the route between the two New Waypoints remained as it was.

This is significantly different behaviour, and there now seems to be no consistency in how the various types of point are treated.

[FrankB]

I'm replying to "Any suggestions about any of the above, or similar experiences gratefully received. I'll phone up Garmin see if I can find a way to send the evidence."
The short answer is: Yes I can confirm that bullets 1 (The moving Shaping Point) and 2 (The Closest Entry Point glitch) have bothered me also.

Read on:
Since I few months I'm the proud owner of a Zumo XT. Before that I had a Zumo 220, so that's quite a step forward. Together with my son I drove parts of the 'Deutsche Motorrad Strasse West', wich is a really nice route. However on 2 occasions I missed the route;
1. Because I had changed a via-point to a shaping-point on the XT. It lead to moving the point off of the original route. I think this is a bug, but not too bad, one can just 'not touch the route in the unit'. And you're OK.
2. I started a route, by selecting 'closest entry point' while not being on the route. The XT wanted me to take > 50Km over the 'autobahn'.
Investigating further, at home I found:
- If you're on the original route, it works better, but not always. Moving a few kms further on the route can make a difference.
- Setting 'Recalculate' to OFF does not have an effect.
- A reliable, but time-consuming method, is to create a track from the route in Basecamp. Send that to XT. On the XT convert the track to a trip. This method is also known as the 'Mrgps method'.
- Until there is a better method, hopefully a fix from Garmin, I decided to also send the track to the XT. In the XT i check 'Always show on the map' and give it a color (blue). That way I immediately notice when the XT chooses a different route.

The routes where imported and modified in Basecamp, containing only 'Alerting points' for hotels.
Zumo: Firmware 2.80 when I noticed the behaviour, 2.90 hasn't made a noticable difference
PC: Windows 7, Basecamp 4.7.2
Map (Zumo & Basecamp): City Navigator Europe NTU 2021.1
Profile: Motorcycling/faster time.
Avoid: Unpaved roads, Ferries, U-Turns, Carpool Lanes

[jfheath]

Thanks for that update [mention]FrankB[/mention]

My previous comments were some time back now, and things have moved on a bit.

1 - On the Zumo, Changing a Via Point to a Shaping Point does indeed move the route point and renames it. This has now been acknowledged by Garmin as a software fault and has been sent tot he development team to sort out. The same fault also appears on the 595 software, but since my 595 is actually a 590 with the 595 software, I didn't press this issue. Hopefully, they will also sort out the issue with routing points sent from BAsecamp not retaining the name (unless they were created as Waypoints). I mentioned this, but didn't push it.

2. When skipping a route point with the skip option, All of the sections of the route are recalculated. This is true, and it has been the case since the 590. I just never noticed it before. This doesn't seem to be the case if you just ignore a shaping point and rejoin the original plotted route after a shaping point - the remaining sections of the route reamins intact. (By 'section', I mean the sections that are between Via Points - ie a route with a start, end and 3 other Via Points (St V1 V2 V3 En) will have 4 sections, irresepctive of any shaping points in between.)

I tested this by creating a route with Via Points on the main road and a number of obvious loops off a main road, held in place with shaping points. I then stripped out the shaping points on transfer to the Zumo. This keeps the shape of the original route, including the loops, but only the Via points are retained. Any section of the route that is recalculated by the Zumo will follow the main road and the loops will be lost.

3. Closest Entry. For my tests this seems to be working properly now. Since V2.70. But I had to install one of the upgrades twice. The first half fixed the fault, a couple of days later it found the same software version again and I installed that. After that 'Closest Entry' worked just fine.
The description you gave matches what happened with one of my earlier tests, which it would not repeat after the upgrades.

One of the earlier fixes seemd to select one of the closest Via Points and head for that. Sometimes it would select the closest shaping point - but only if it had been created as a Waypoint with the Basecamp Flag tool, and then changed to a shaping point. This suggested to me that the programmers do not seem to be using the same definitions of the various points that the Garmin help website is using ! By doing this, they seem to have introduced two sorts of Shaping Points. Ones that are ad-hoc and put in place using the route tool or insert tool, and those that have been created as a Waypoint and then converted to a non-alerting point.

Thankfully, this seems to have been corrected. My etsts are that it finds the closest point of the original route to where you are now, and heads for that point. Once you are there it follows the route to the next shaping point or Via Point - whichever is next.

I thought I had tested this very thoroughly - giving it every chance to get it wrong - and I couldn't get it to fail. There may be issues with traffic if you have such things turned on, or the riding profile that you have built up. My Zumo was cleared of all of that before I tested it. More than happy to look at your route to see if I can produce the same issue.

nb - it takes quite a while (weeks) for Zumo to get back to you having contacted them - but they do eventually respond.

[FrankB]

Thanks for your reply.
About changing to a shaping point. Good to hear that Garmin has confirmed this.

About the 'closest entry point glitch'.
My device is on firmware 2.90, and I can still reproduce it. I did a test on Saturday (it was hot!) driving my bike. Sometimes it will take you to the shortest point on the route, sometimes straight to the end-point. This is nearly impossible for me to demonstrate.

Another way to demonstrate the behaviour is on the device with GPS OFF, and then simulate the route. (You are probably familiar with this method)

Settings/Navigation/GPS Simulator.
View map. Move to the point of interest. Click on it. A small blue circle appears. Click on the description bar. Scroll down to 'Set Location'.
Now if I start a trip, I get the same results as 'in real life'.

I'm happy to post my route. See the attached zip. You'll find:
- The route from Neustadt to L512. It has some shaping points, and a via point. (The Hotel)
- The converted track from this route. Once loaded in the XT, I set it to 'Show on Map', color Blue. The only purpose is to quickly see if the route is correct once started.
- 3 Waypoints. Showing the effect of the starting location on the route.
- Screenshot of the tests that I did.

It is worth mentioning that when we actually drove the route there was a detour. So it was not possible for use to get to the starting point. Else all would have gone unnoticed.

Let me know if you need anything else. Thanks in advance

[jfheath]

Hi

Thanks for that information - very clear.

My Zumo is set up exactly the same - same ,aps, latest software 2.90, maps 2021.1 - in fact the route loaded directly into my Zumo without needing to recalculate. Turned GPS off, loaded your track, checked the route loaded ok, set my position to somewhere nearby. I got exactly the same results as you. It ignored all of your shaping points and found the quickest way to the nearest VIA point - near the end of the route.

Now this is what it used to do. I'll need to go back to my notes and find out when they fixed this problem. But they did fix it.

Now your route list has a start point, an end point and ONE Via Point. All of the rest of them are Shaping Points. These are the ones that show up as little blue circles when you first load your route. The Via Points show up as flags.

Forget what it says in Basecamp, telling you how many Via Points you have. It is lying. A Via Point is an alerting route point. A Shaping point is a non alerting route point.

If you change ALL of your shaping points to Via Points the closest entry point works after a fashion.

What the original software did was treat a Waypoint in the route differently. ie one that had been created with the flag tool in basecamp. These become Via Points by default. If you then made them into shaping point (non alerting) - the route would look exactly like yours - with little blue dots, and the closest entry point would then work.

That was the original software. I haven't had time to try the same trick on your route - but I will. It was updated a couple of times (arounf 2.6 or 2.7 - I need to look it up), and after that, it didn't head for a point at all - it found the closest point on the route. I had got a complaint in about it at the time, and withdrew it because the fault had been fixed.

It is now back as it was again. I will resurrect the complaint. This is ridiculous. Ive updated to 2.8 and 2.9 since I last played with this, so I don't know when the fault re-appeared. All of those Shaping points have just been wiped out. Just like it would if you selected the End point as the next destination.

Bottom line is - Yes I have got the same results. Yes it certainly appears to be wrong. They seemed to have fixed it, but it now seems to have gone back to the same issue it had in the first place.

[FrankB]

First of all. Thank you very much for taking the time to dig into this issue.
Second: I'm really pleased that you come to the same conclusions as I did. I understand and agree to all of your remarks.
If you want more information, dont hesitate to post your question.

[jfheath]

I've just dug out my old correspondance where I reported the fault - way back in V2.50 From my records 2.6 didn't fix it. 2.70 improved it and a couple of days later another 2.70 made it work really well.
No - Oops - that's wrong. Prior to 2.4, Closest Entry was behaving very oddly. I have the notes I made at the time. V 2.5 improved it a lot, and then when I connected a couple of days later there was another 2.5 software update.
Subsequent tests with 2.5 and (i think 2.6) and it worked very well, locating the closest point on the magenta line and following the route from there, including all Via and all Shaping points - no matter whether they started life as Waypoints or not.

I've just written back to them using the same reference saying that it is broken again. Might be worth other contacts too. I found it easier to contact by phone - briefly tell them the problem and get a reference number.

Have you tried a circular route ? With the end close to where you are ? It makes for a very short ride - the end point is the closest Via Point, and it heads for that. It seems to be locating the Via Point that is closest to you - and then behaves just like when you select a Via Point from 'Next Destination' Goes straight to it ignoring any routing points in between.