Calculation changes Route

[YouHa]

Hello,
maybe someone has an idea...

I make a GPX Route (Basecamp or GoogleMaps) import that to Zumo XT and open the route in saved routes.
Now the map shows the correct route, but when I "start" the route and it gets calculate, the result is different
from the planed route.

Thank you

[jfheath]

"Calculation Changes Route".

Your title sums it up in a nutshell - Yes it does.

The reason for this is that the XT uses different algorithms for calculating routes, the details of which many of us are still tryiing to work out. Whatever it is actually doing, it displays very different behaviour from the Zumo 590, 595, 39x series, 66x series and 550. Once we understand exctly what it does, we will be more able to plan routes that take into account how it behaves - so that it doesn not calculate a different route.

Let me give an example. Before the XT came along, 'faster route' gives a more predictable result. Motorcyclists don't normally want to spend their time on a motorway / autobahn, so opt to take the old main roads or the side roads that head towards the same end point. But we know very well that given half a chance, the Zumo will calculate a route that uses the fast, multi-lane motorway. So we predict this behaviour and use route points (shaping points or via points) to convince it that this isn't where we want to go.

The XT seems to do something similar with other types of route. In the UK we have A roads and B roads. Some A roads are major trunk roads. I am guessing that some roads are classed as being faster than other kinds of roads. So - perhaps Major trunk A roads are classsed as fastest, A roads are fast, B roads are not so fast. The routing seems to work out long routes just to include A roads. I have noticed in a few experiments that if you try to tempt the XT to take a 'short cut' B road, it does not take the B road.

I have also noticed that if you transfer a route from Basecamp, or a gpx v1.2 route from MyRouteApp then as always, the XT receives the route exactly as it was planned. If you force the XT to recalculate the route - eg by changing vehicle or changing avoidance settings, or skipp a route point, then the entire route gets recalculated.

It used to be that if you deviated from the plotted route, the route would be recalculated. Even on some early tests that I carried out on the XT, this was the case.

But yesterday I was out with the XT in the car, and I followed a Basecamp route on the way to my destination and an MRA route on the way back. In both cases I deviated from the original route and expected the XT to find a new way to the next route point. It didn't. What it did instead was calculate a route back to where I had left the original route. The original route stayed put. I have observed this in certain circumsatnces before, but never when just deviating from the route. As soon as you join the route again, it continues as before and I think it does this even if you have passed Via Points - but I need to check that.

It seems to be now placing much more importance on following the original line - almost like following a track - than it does on getting you from one route point to the next. But yesterday, the route that it had one the screen never changed. And yes - autorecalc was turned on, and I was following a normal trip.

Now that IS new behaviour from my previous tests. I just don't know precisely when the change in software behaviour happened. It is also different from what is described should happen when you deviate from a route. And - it makes some of my notes I made in the 'Everything that you need to know' pages on this site, very wrong.

If what I have observed is correct, the XT is becoming a different beast from the one that I thought that I had bought. I can see advantages and disadvanatges, but one I get used to what it is doing, I will be able to work with it.

Ping @Peobody, @Fxwheels, Ping @rbentnail, @lkraus

[rbentnail]

But yesterday I was out with the XT in the car, and I followed a Basecamp route on the way to my destination and an MRA route on the way back. In both cases I deviated from the original route and expected the XT to find a new way to the next route point. It didn't. What it did instead was calculate a route back to where I had left the original route. The original route stayed put. I have observed this in certain circumsatnces before, but never when just deviating from the route. As soon as you join the route again, it continues as before and I think it does this even if you have passed Via Points - but I need to check that.

It seems to be now placing much more importance on following the original line - almost like following a track - than it does on getting you from one route point to the next. But yesterday, the route that it had one the screen never changed. And yes - autorecalc was turned on, and I was following a normal trip.

Now that IS new behaviour from my previous tests. I just don't know precisely when the change in software behaviour happened. It is also different from what is described should happen when you deviate from a route. And - it makes some of my notes I made in the 'Everything that you need to know' pages on this site, very wrong.

If what I have observed is correct, the XT is becoming a different beast from the one that I thought that I had bought. I can see advantages and disadvanatges, but one I get used to what it is doing, I will be able to work with it.

Ping @Peobody, @Fxwheels, Ping @rbentnail, @lkraus

Yes, the bold text in your quote is what I am seeing. Even with Auto Recalculate on, the XT tries to get me to u-turn and backtrack to get to the point I exited the route rather than take me straight ahead to return to the route. But in contrast, I did not experience the XT "rejoining" the route when I was back on it. I have never seen this behavior, my XT has yet to do this. It ALWAYS wants me to turn around to visit some phantom point behind me. And very often when I do not follow its "Make A U-Turn" instruction, it says "Cannot Calculate Route" and closes the route itself. I then have to restart the route. Occasionally this happens more than once when the route has already been rejoined- it still wants me to go backwards to some point I'm already well past.

Very confusing behavior indeed.

[jfheath]

To be sure about what I saw yesterday, I need to revisit a number of tests that I did over a year ago.
More recently I observed this behaviour - but only after I had skipped a previous point, which resulted in the XT recalculating the whole of the original route and choosing roads which meet its definition of 'faster'. Then when I ignored its route, it kept trying to take me back.

I don't know for certain whether the behaviour has changed or whether the circumstances were different.

Since you had issues with Vias, I'll check that out and make sure that my observations were correct.

[Fxwheels]

I haven't yet experienced this problem above, but I'll do some tests and report back on it. In my experience if I deviate from the route, it normally would recalculate and rejoin at some later point. Occasionally it wants me to U-turn, but the further I go it recalculates and finally brings me on route.
A problem can appear when restarting from the "closest entry" it will take you back. But I always have a track running along. If the navigation behavior upsets me, I just ex-out the route and follow the track.

[rbentnail]

I'm beginning to think that "closest entry point" actually means "closest via point".

[Fxwheels]

Well, in the older version (not sure which) it used to be a nearest via or a shaping point. Now they suppository changed it to a nearest magenta line. This function was working flawlessly in the old Nuvi 1450/1490 (~year 2012). Of course the entire software was much more simple. After they redesigned the software to be more sophisticated they "created" more bugs in it along with it. Will the bugs are going to be ironed out?... I wouldn't hold my breath.

[Peobody]

When it happened to me, my sense was that it was insisting that I return to the point where I deviated from the route. I had one ad-hoc shaping point left to pass through in the original route so at first I thought it was trying to get me back to it but it didn't reroute me to that point like I would have expected. I then skipped that point but that didn't make any difference. The XT continued to try to u-turn me back the way I had come.

[Oop North John]

When it happened to me, my sense was that it was insisting that I return to the point where I deviated from the route. I had one ad-hoc shaping point left to pass through in the original route so at first I thought it was trying to get me back to it but it didn't reroute me to that point like I would have expected. I then skipped that point but that didn't make any difference. The XT continued to try to u-turn me back the way I had come.

The Streetpilot 3 would tie itself in knots, and be useless for miles until the developers added an algorithm for it to fully re-calculate the route after the operator ignored the suggestions three times. Wouldn't it be nice if they could resurrect that logic from 15 odd years ago on their present flagship GPS

[jfheath]

I did a lot of testing on CEP and it used to head to the nearest Via Point, then to the nearest Via Point or Waypoint (ie saved point), then to the nearest Via or Shaping point (reagrdless of anything being a waypoint). Then with v 2.90 is really did go to the closest point on the route.

I don't know if that has changed, but whwnever I have used it, it certainly doesn't head for any route point specifically.

Today I used the XT to plote a route. A destination. That was all.
Then I deviated from the route. And it worked flawlessley. When I deviated, it suggested a new way pretty quickly - initially this was to turn round and go back, but then it found other ways, going ahead. Just like my 590 does.

So then on the way back I added another point. Initially a Via (becasue these are easier to add), but I changed it to a shaping point. The shaping point did not show on the map, but the route stayed put.

The I tried a route with a Via and I used the shaping tool to add a shaping point. That too worked OK, just as I expected.

I am wondering if Garmin have had so many complaints about the route being altered from what was created in Basecamp that they have done something to address that.

MRA gpx v1.2 sends no shaping points to the XT. Yet the route stays put - even if you deviate from it. The XT will not recalculate it.
It turns out that routes from basecamp do the same thing, but BC routes contain the shaping points (although you can strip them on transfer).

So my current working theories are these (and I put it in a box because these are hypotheses (null hypotheses) which need to be tested.

Theory 1: If you send a route that is plotted on roads, then the XT does not currently recalculate it. and it will always try to get you to go back if you deviate.

Theory 2: If you send it a direct route - ie no roads plotted, just the shaping and via points. The XT will work out the route and it will behave like the 590/660/550 used to. If you deviate, it will find a new way to get you to the next route point.

Theory 3: If you force a recalculation of a route with plotted roads (ie with the ghost points). Then it recalculates a new route, but treats it like that is the route that you plotted and will stick to it.

Theory 4: Garmin have changed the routing software since I last tested it.

If any of the above turn out to be true, then I have a lot of re-writing to do. It will make my essay in "Everything that you need to know" a load of boll-twoddle.
Please dont read the above as fact. I've seen al l of these happen.