Routing errors (again) - any idea of a point of contact that may care ?

Having Garmin zumo XT problems? there is loads of help and advice in this forum
jfheath
Posts: 2199
Joined: Sat Oct 19, 2019 4:17 pm
Location: West Yorkshire, Uk
Has liked: 267 times
Been liked: 587 times
Great Britain

Re: Routing errors (again) - any idea of a point of contact that may care ?

Post by jfheath »

Thanks for that. The reason I asked is that I noticed that the caption underneath said the same thing no matter what the 3rd screen radio button said. Also that changing the settings when in Bike mode also changed the settings in car mode. Which is definitely wrong.
However, when I did a factory reset, everything was back as it should be.

I noticed when I was having problems with music and sound, that it only happened after I had been fiddling with the Explore website. After doing a reset, it behaved normally. I repeated this a number of times - each time I connected to Explore for the first time, the sound would be on off on off very rapidly. Sometimes it played v e r y s l o w l y.

I'm wondering if the same is happening with the Avoid Tolls.

I did a few tests. But I have to redo them as I forgot that doing a reset ticks all of myMaps - and I forgot to untick my OSM maps, which will probably have affected the results.

I tested your routes. My version of your routes created in BC and transferred (but with a correct positioning of Point B), and routes created on the XT Trip Planner using favourites with A-C, B-C, A-B-C, B2-C, A-B2-C.

Every version of each test behaved exactly the same as the every other version of the same test - both with Tolls allowed and tolls avoided. The inaccurate positioning of B makes no difference. The transfer from BAsecamp makes no difference. I even tried A to shaping point at B to C in case it was to do with Garmin's waypoints. So the XT version, the SN version and the JH version of each test produced identical results.

It seems that something is making it avoid certain roads around the are of Point B. So I got onto google street view, and looked at the roads. There are areas there that in 2015, the roads around Hambledon were closed.

So I did a simulation.

Try simulating the route. There are no speed limit signs on the Zumo screen - except near the A3 - but the simulation normally drives up to the speed limit on open roads. Watch what speed it sticks to. That might be a reason why it goes for main roads if it thinks it can do only 30-40mph on most of the rest of it. It doesn't explain it fully, but its getting closer.

I need to double check my results for the tests without OSM being ticked.

Today, I discovered something really important. I found another way that doesn't work.
Scottnet
Posts: 77
Joined: Thu Jun 04, 2020 4:57 am
Has liked: 1 time
Been liked: 6 times

Re: Routing errors (again) - any idea of a point of contact that may care ?

Post by Scottnet »

Thanks for the update. I suspect the truth is simply a software bug.

A to c works
A to b to c does not.

Looking forward to your next update
jfheath
Posts: 2199
Joined: Sat Oct 19, 2019 4:17 pm
Location: West Yorkshire, Uk
Has liked: 267 times
Been liked: 587 times
Great Britain

Re: Routing errors (again) - any idea of a point of contact that may care ?

Post by jfheath »

Not quite.

If you set the XT to avoid toll roads, A-C works. The other two tests take a sweep north to Petersfield
If you set the XT to allow toll roads, A-C takes a sweep SE first, then NE. The other two work.

Avoid Tolls Set

Toll JH01a Avoid A-C .png
Toll JH01a Avoid A-C .png (77.41 KiB) Viewed 1141 times
Toll JH02a Avoid B2-C.png
Toll JH02a Avoid B2-C.png (90.46 KiB) Viewed 1141 times
Toll JH03a Avoid A-B2-C.png
Toll JH03a Avoid A-B2-C.png (80.84 KiB) Viewed 1141 times
Allow Tolls Set

Toll JH01a Allow A-C.png
Toll JH01a Allow A-C.png (79.23 KiB) Viewed 1141 times
Toll JH02a Allow B2-C.png
Toll JH02a Allow B2-C.png (87.95 KiB) Viewed 1141 times
Toll JH03a Allow A-B2-C.png
Toll JH03a Allow A-B2-C.png (77.62 KiB) Viewed 1141 times

Im sure the tolls thing is a red herring. I'm wondering about the affect of the presence of B2 in the route. But so far, I cannot see a clear pattern

Today, I discovered something really important. I found another way that doesn't work.
Scottnet
Posts: 77
Joined: Thu Jun 04, 2020 4:57 am
Has liked: 1 time
Been liked: 6 times

Re: Routing errors (again) - any idea of a point of contact that may care ?

Post by Scottnet »

I agree that the tolls setting wirednessis is complicating things. I suggested we focus on why,

When B is a shaping point it works as expected but when it is waypoint it does not.

I think that narrows does the scope of the question and the permutations. The developers should be able to debug based on that.

Should I raise this as a new question to Garmin ?
I will ignore anything tolls related.
jfheath
Posts: 2199
Joined: Sat Oct 19, 2019 4:17 pm
Location: West Yorkshire, Uk
Has liked: 267 times
Been liked: 587 times
Great Britain

Re: Routing errors (again) - any idea of a point of contact that may care ?

Post by jfheath »

Yes - how will you do it - by email or by phone ?

I think that Via Point / shaping point issue is to do with the fact that it calculates the routes between via Points. But that is a point to mention.

I have just been playing this morning with some variations. Some changes. My unit is now keeping the toll setting with the vehicle, and it is keeping the vehicle and faster time / shorter distance with the route. So that is correct behaviour. As a result, whereas every route chnaged depending on whether or not tolls were allowed, I know have (I think) two routes which give the correct routing. I've set Avoid tolls for the car and allow tolls for the bike, and instead of going back to navigation each time, I'm just changing the vehicle in the route settings.

However. The whatever I do, the shortes route never comes out the shortest route. Faster time sometimes doesn't come out as faster time, and I have repeated that on roads around here.

I am suspecting that the algorithm uses the trafficTrends data that comes with the maps, but there is no option to turn it off on the 595 or the XT. At some point I'll switch my 595 to be 590 and see if that reproduces the same 'fault' if I allow it to use trafficTrends.

But for the Garmin conversation.
Faster time doesn't produce faster time.
Shortest route doesn't produce shortest route.
It isn't a mapping problem - it does the same with OSM maps. It does not do it on the 595 witht he same maps.
And changing some settings changes the route (eg toll roads, when there are no toll roads around).

I'll do a bit more testing and get in touch as well. I like to get past the standard do this / do that / and then 3 emails later I can have a proper conversation. If I make it clear that I have done all of the usual stuff, then we can start from there.

Today, I discovered something really important. I found another way that doesn't work.
sussamb
Posts: 1713
Joined: Mon Apr 23, 2018 1:06 pm
Has liked: 306 times
Been liked: 353 times
Great Britain

Re: Routing errors (again) - any idea of a point of contact that may care ?

Post by sussamb »

Fascinating what you've already discovered so far. Watching with interest :D
Scottnet
Posts: 77
Joined: Thu Jun 04, 2020 4:57 am
Has liked: 1 time
Been liked: 6 times

Re: Routing errors (again) - any idea of a point of contact that may care ?

Post by Scottnet »

I will raise by email. I already had but I got the std answer to report it as a mapping issue.

Keep me posted on responses you get from your email.
jfheath
Posts: 2199
Joined: Sat Oct 19, 2019 4:17 pm
Location: West Yorkshire, Uk
Has liked: 267 times
Been liked: 587 times
Great Britain

Re: Routing errors (again) - any idea of a point of contact that may care ?

Post by jfheath »

Got a reply within an hour by email. This was sent by someone or something that skims the message for key words and doesn't take much notice of the rest of it.

This is normal in my experience. I suspect it may be an automated response.

So it presented 3 videos to look at. How to update maps, how to update software and one which was slightly more informative dealing with how different circumstances can affect routing, and how car and motorcycle setting will choose different routes. Nothing that I hadn't already indicated that I had considered. I have yet to find a situation where car and motorcycle choose different routes if the settings are identical. But if the settings are different, then of course they will.

I try hard not to feel patronised by this. But its a process akin to a triage when you go into a hospital.

But I now have a reference number, so if the next reply doesn't help, I can talk to someone on the phone.

Today, I discovered something really important. I found another way that doesn't work.
Scottnet
Posts: 77
Joined: Thu Jun 04, 2020 4:57 am
Has liked: 1 time
Been liked: 6 times

Re: Routing errors (again) - any idea of a point of contact that may care ?

Post by Scottnet »

Have you made any progress on this?
jfheath
Posts: 2199
Joined: Sat Oct 19, 2019 4:17 pm
Location: West Yorkshire, Uk
Has liked: 267 times
Been liked: 587 times
Great Britain

Re: Routing errors (again) - any idea of a point of contact that may care ?

Post by jfheath »

Sorry

Not really. I had two really stupid email exchanges where it was obvious that the person responding hadn't bothered reading the content of the email.
Gave me a reference page 'cos the content of the videos on there had a word that matched one of the words in my first sentence. Basically telling me how to suck eggs. But he doesn't know how much I know, so I sort of expected that.

The second one he was asking me stupid questions like was I using circular routing ??? Again he hadn't read the email - only the title.

I said we weren't going to get any further like this, could I have a phone call with someone. So he gave me a number.

But what I did get out of it was a reference number, so when I talk to someone, they can pick up the first email.

That was late Monday afternoon. I've not had time to make the phone call yet. Very busy week. I need a clear slot with no possible interuptions.

I'll have a chat with them next week.

Today, I discovered something really important. I found another way that doesn't work.
Post Reply