Weird Routing Behaviour (2)

[guiri]

You are using MRA ? Do you have the gold version ? Are you using HERE maps ?

Yes, I have MRA Gold and used the HERE routing to plan the trip.

I exported the route and track with gpx 1.1 from MRA and sent the file to the XT using Garmin Drive (no basecamp involved). Once on the XT, I made sure that the track was visible (shown as default blue in my XT and I'd made it maximum width). In the Trip Planner, I then selected the route and hit Go, and selected next destination as my first via point.

I followed this process a couple of times adding shaping points as required. In the Trip Planner, the calculated route (in magenta) did not initially match my track (in blue). One of the reasons was a difference in the HERE mapping. MRA has a correctly closed road, whereas the latest version of the mapping on the XT now has the road incorrectly as open. I used a shaping point to avoid the XT trying to navigate that closed section.

[jfheath]

Thanks for the update @guiri.

The question was important because your answer eliminates a few possible causes. I'm in the middle of writing my next report for Garmin Support - after the Xmas break, so I'll come back to this later.

Just one quick comment. MRA's GPX 1.1 does indeed contain all of the Via Points and Shaping points (unlike 1.2). However, it does not contain any route information. All that MRA sends is the route as a series of routing points - in the correct order and with the correct Via/Shaping point identifiers. But no route. The track contains the 'plot' of the route that was created on MRA. But the route definition does not.

So with MRA's GPX v1.1 - the XT will always need to calculate the route using the route point information and it will do that in its own unique way. So it won't be just the fact that you have 'roads not yet built' to contend with. The XT will probably always come up with a different route from the one created in MRA.

That isn't a bad thing as any deviation, skip, use of closest entry point, traffic issues - will also cause the XT to recalculate - so I think it is good to find out what route the XT is going to produce before you set off. And the superimposed (underimposed ??) track on the same screen helps to inform you of the changes to the original.

The XT builds up a picture of your riding history, and uses that to adjust the routes that it generates. I suspect that it also uses historic traffic data that is built into the maps (but I don't know). I have no other way of explaining that when I plot a route on one day on the XT it does something different on another day. I've only had my new unti a few weeks, and it has been idle over Christmas, but already I have performed a system reset to get rid of any profiles that it might have built up.

[guiri]

Just one quick comment. MRA's GPX 1.1 does indeed contain all of the Via Points and Shaping points (unlike 1.2). However, it does not contain any route information. All that MRA sends is the route as a series of routing points - in the correct order and with the correct Via/Shaping point identifiers. But no route. The track contains the 'plot' of the route that was created on MRA. But the route definition does not.

It's good to be clear on the semantics! Whilst MRA describe the 1.1 export as being for "route, track and waypoints", I will refer to it as bringing over the routing points and track in future. Even after adding some extra shaping points, the XT still took a different route than my (MRA) track at one point, which was a sneaky rat run down a side street to avoid some traffic lights (so nice one XT)!

By the way, do you know if the XT software team is based out of Calgary?

Cheers
guiri

[jfheath]

By the way, do you know if the XT software team is based out of Calgary?

I haven't got a clue. I don't have anything to do with Garmin - other than the fact that I want my XT to work as I expect, so any issues like this one - I want to give them as much quality information as I can to get it sorted.

The MRA v1.2 format where shaping points are missed out but the route is retained - a similar option is also available in Basecamp. You can ask it to strip out the shaping points, and providing the route is not recalculated byt the XT, it will follow your route exactly. I tested the MRA 1.2 routes on just one occasion to arrive at that conclusion - so it might need further research, but once was good enough to see what it did. But MRA has a BIG plus point as far as I am concerned - the names of route points are not changed in the transfer process which is a big issue with Garmin and the XT.

[guiri]

By the way, do you know if the XT software team is based out of Calgary?

The MRA v1.1 format where shaping points are missed out but the route is retained - a similar option is also available in Basecamp.

Did you mean MRA 1.2? In which case I tested it to discover that it looked good in the Trip Planner until you hit Go, at which point it recalculated and did its own (fastest) thing between the via points. For my ride yesterday I then decided to use MRA 1.1 and could at least see the effect of the shaping points in the route calculation that the XT was making.

I'll try and do some simple tests on my commute (which I do on foot so at least safe to play with the screens)!

[jfheath]

Did you mean MRA 1.2?

Yes - I did mean MRA gpx v1.2. That's what I said in an earlier post. Goodness knows why I said 1.1 in the last one.

Thanks for spotting that error. I've corrected my post. Its Ok to leave it in the quote, otherwise these following two posts don't make sense !

[jfheath]

Right coming back to your sequence that you described a few posts back.

I think the behaviour that you observed may be quite normal. It is perfectly acceptable for the XT to use its faste route algorithm and choose to take you BACK to a motorway junction to route you down a very much faster road. It doesn't mean (necessarily) that it is stuck in a Repeated U Turn loop.
It might be, but in that situation I would assume that it is simply heading for the motorway. It will do this until the way you are riding beomes faster.

If it continues to do this, then there is an issue - but not with your particular unit. What seems to trigger it is if you skip route points or choose closest entry point and then deviate from the plotted route. But I have yet to see it happen if the route is created on the XT. I assumed that it wouldn't happen with MRA routes, but your scenario may seem to contradict this notion.

But you have two odd circumstances in your particular example. One is the presence of the motorway which may be legitimately re-routing you. The other is the map showing the closed road as being open.

When out and about, 'Skip' will show you the next point to be skipped. Edit Route - which can be added to the right hand buttons on the map screen, and is also available fromt he 3 dot menu bottom right - will show you the next Via and the next shaping, point . It does not tell you which comes next - but you can determine that from the Skip button.

I'm going to do some testing with MRA routes to see if the same RUT (Repeated U Turn) behaviour shows up in circumstances similar to my other tests. I think that it will not, but time will tell. I'll feed back to you when I know.

[Oop North John]

I'm not sure if it's been mentioned before but has anyone else noted that if the end and via points are already in the XT, and you get the XT to calculate it, then if you don't follow it it won't do the U-Turn madness, but behaves as you'd hope / expect?

The same route imported from Basecamp = U--Turn madness.

[Peobody]

Hmmm... When you import a Basecamp route its waypoints are added to the XT. The U-turn madness occurs following a mid-route recalc so you would think that it would produce that same result as if the points had already been on the XT.

[jfheath]

Thanks @Oop North John . Yes, it was one of the things that I was able to confirm about the behaviour to tech support, just before Chritamas.

I'm currently working on trying to find out exaclty which part of the BAsecamp route causes this. If indeed it is the BA=asecamp route. I am suspecting the SubClass string and have a series of tests set up (ready for the ice to disappear) to test how routes behave if the Subclass is set to default ....ffff0000.... string, and if it is ommitted at all.
Also what happens if a direct route profile is sent from Basecamp.

Forcing a BC route to recalculate after it has been received by the XT doesn't cure the problem. But a route created on the XT using favourites/waypoints that have coem from Basecamp does seem to slove the problem and the route behaves normally.

I reported the fault about 2 years ago, and it led nowhere. This time the fault was more visible - when it gets stuck in the U turn madness, the track log becomes fragmented, starting a new section at each demand for a U-turn. I've demonstrated that the two issues are related, and they have become more interested in the repeated U turn situation.

I'm talking with @FrankB who is also looking into the Subclass string - again to see if something in that can affect the nature of the route points.

@Peobody - Your repsonse came in as I was typing the above. Routes created on the XT seem to be OK. Routes created on the XT screen with the Waypoints that were transmitted from BC (and which are stored in Favourites / Saved), also seem to behave themselves.
But as mentioned above - a route from BAsecamp which is then recalculated by the XT seems to not behave itself in the circumstances described above. But more tests to carry out.