Ongoing routing problems

[scuba_al]

Ive been a TomTom user for over a decade.
For 5+ years I have been reporting significant routing errors in Australia that never get corrected.
When you choose route between A and B and TomTom choose a route 40km and 15 mins longer than Google and Apple.
If you add a mid-point stop to the route TomTom suddenly finds it and shows it as 40kms and 15 mins shorter!
Its not the app - its the underlying algorithm because its exactly the same in the mydrive web app.
I have reported so many time to try and help resolve it but never any change.
LAst June TomTom gave me a 'free' 6 month extension while they 'resolved the issue'.
Needless to say nothing has happened.

If anyone wants to see the problem use MyDrive to navigate between Canberra (ACT) and Huskisson (NSW).
Tomtom recommends fastest route as 3h5mins....which it isnt.
Now add Oallen (South Coast) to the route.
It suddenly drops to 2h 49 mins.
Same result on every platform so its the algorithm NOT the application.
I am sick of waiting.

 

[canderson]

I doubt that it's either the algorithm or the application that's at the heart of the matter, though one can argue about the way the algorithm treats tertiary (actually, what's 4th? quaternary?) roads when creating routes. Irrespective of any speed limits, both actual traffic data and road class play parts in producing the route.


https://plan.tomtom.com produces the following:

maps.google.com produces this:

I suspect that this is because one of the many small roads involved in the Google route is classed in the TomTom map database as something just this side of a goat track (and to be fair, a road level view shows a few spots to be not much more than that in places), and will have assigned to it a large penalty for doing initial routing using it. It would take quite a bit of time to track down the precise section that is causing this since there are a great many segments involved in making that W>E crossing, but I'd bet my next paycheck that it's a 'road class' issue. Have seen these before. Once you get past the area(s) at issue, the device suddenly realizes that the rest of the journey is 'downhill' and shows the reduced time.

I had one of these not far from my home ... 4 lane, divided in many places, 45mph (75km/h) speed and only a few traffic lights. The first part of it had been incorrectly categorized as one of the lowest of road classes, so devices (and the online planner) never preferred it over other, parallel choices. Only after driving past the problem area did the device quit trying to reroute me to a parallel road and would recalculate and discover that life wasn't so bad after all.


-----------------------
Edit: Without reviewing the entire route, I can confirm that the Google route includes at least some Net 2 Class 5 road segments. If I searched long enough, my nightmare would be to find a Class 6 road in there anywhere.

If you take the route, at what location does the TomTom give in and plan something like the Google route for the remainder of the journey?

 

[Niall]

Try changing the start and/or endpoints of this route by a couple of hundred meters and you add an hour to the journey

Plan.TomTom.com

TomTom Route Planner

plan.tomtom.com

 

[canderson]

Interesting experiment, Niall! Evidently you moved so as to make the goat track route, even with its time penalties, better than the longer route. Guess it depends upon how close you are on start point to the B52?

 

[scuba_al]

I doubt that it's either the algorithm or the application that's at the heart of the matter, though one can argue about the way the algorithm treats tertiary (actually, what's 4th? quaternary?) roads when creating routes. Irrespective of any speed limits, both actual traffic data and road class play parts in producing the route.


https://plan.tomtom.com produces the following:

View attachment 6120

maps.google.com produces this:

View attachment 6121

I suspect that this is because one of the many small roads involved in the Google route is classed in the TomTom map database as something just this side of a goat track (and to be fair, a road level view shows a few spots to be not much more than that in places), and will have assigned to it a large penalty for doing initial routing using it. It would take quite a bit of time to track down the precise section that is causing this since there are a great many segments involved in making that W>E crossing, but I'd bet my next paycheck that it's a 'road class' issue. Have seen these before. Once you get past the area(s) at issue, the device suddenly realizes that the rest of the journey is 'downhill' and shows the reduced time.

I had one of these not far from my home ... 4 lane, divided in many places, 45mph (75km/h) speed and only a few traffic lights. The first part of it had been incorrectly categorized as one of the lowest of road classes, so devices (and the online planner) never preferred it over other, parallel choices. Only after driving past the problem area did the device quit trying to reroute me to a parallel road and would recalculate and discover that life wasn't so bad after all.


-----------------------
Edit: Without reviewing the entire route, I can confirm that the Google route includes at least some Net 2 Class 5 road segments. If I searched long enough, my nightmare would be to find a Class 6 road in there anywhere.

If you take the route, at what location does the TomTom give in and plan something like the Google route for the remainder of the journey?

thanks for your comprehensive reply. I know its the way the algorythm interprets the costing applied to the various roads and 5 years ago one of the raods was unsealed (in your terms a goat track). for over 4 years that has NOT been the case and it is bitumen all the way and a well used raod. I have reported this to tomTom over 20 times - including the EXACT coordinates at which Tomtom decides to suddenly use the faster and shorter route = which indicates where the error/miscategorisation is. The frustration is that despite everything I have done to provide all the relevent informaiton to TomTom to have it fixed - nothing has happened. In the meantime all the other map providers have correctly upgraded the road and show the fastest route. I have now dropped my TomTom subscription and am using Google maps for free like everyone else. Very disappointing.

 

[scuba_al]

Interesting experiment, Niall! Evidently you moved so as to make the goat track route, even with its time penalties, better than the longer route. Guess it depends upon how close you are on start point to the B52?

as I said the routing is stuffed

 

[canderson]

thanks for your comprehensive reply. I know its the way the algorythm interprets the costing applied to the various roads and 5 years ago one of the raods was unsealed (in your terms a goat track). for over 4 years that has NOT been the case and it is bitumen all the way and a well used raod. I have reported this to tomTom over 20 times - including the EXACT coordinates at which Tomtom decides to suddenly use the faster and shorter route = which indicates where the error/miscategorisation is. The frustration is that despite everything I have done to provide all the relevent informaiton to TomTom to have it fixed - nothing has happened. In the meantime all the other map providers have correctly upgraded the road and show the fastest route. I have now dropped my TomTom subscription and am using Google maps for free like everyone else. Very disappointing.

If you can supply me with the coordinate data for where the routing changes its mind and the road you suspect is at fault, I'll do my best to see about getting any necessary change made.