Routing quality

[doodah]

I’m interested in hearing from people who have other auto-routing Garmin units in addition to their iQue (GPS V, SPIII, etc). Have you noticed that the iQue often creates lower quality routes?

I have run several experiments where the iQue will compute a route that I would never consider taking, usually because it's longer. Using vias, I force it to use a more rational route and then compare the time estimates. In all cases, my forced route comes out better. When I compute the same route with Mapsource or my GPS V, it comes up with the same route that I had to force into the iQue. From these experiments, it appears to me that the iQue algorithm gives more priority to higher level highways in order to avoid what it thinks may be roads with stop signs, traffic lights, etc. One would think that Garmin would tend to use the same algorithm among all auto-routing products rather than reinvent the wheel just for the iQue.

[Peztone]

Have you tried altering the preferences from Fastest to Shortest or the other way round?

[jonasolof]

A person 850 will tell you that it'll improve with firmware 3

[doodah]

Originally posted by Peztone
Have you tried altering the preferences from Fastest to Shortest or the other way round? Yes I have. And that's what's so confusing! How can the computed "shortest" route be faster than the computed "fastest" route? That seems counterintuitive to me.

[terrier]

Originally posted by doodah
Yes I have. And that's what's so confusing! How can the computed "shortest" route be faster than the computed "fastest" route? That seems counterintuitive to me.

Roads have attributes associated with including an average speed. A route that uses a freeway that is twice as long with a speed 3 times of a backroads route will still get you to your destination faster.

If you want the short scenic route, it would take you there a little bit slower, but sometimes that's not so bad after a long day at work, especially if the day is nice and you have a convertible!:cool:

[tusker]

I have the 'quickest route' preference selected.

Sometimes when I elect not to follow the chosen route, I note the recalculated route often shows an ETA several minutes earlier than Gladys had been predicting on her earlier choice of route. The question that therefore springs to mind is - why doesn't Gladys constantly look for better alternatives (or at least have this as a selectable option) - when on a route??

[rwsmith123]

The problem that doodah was describing is that the calculated shortest route is faster than the calculated faster route. These are calculations not actual driving times, so the shortest route shouldn't be faster than the fastest route.

I've seen the reverse problem where the fastest route was shorter than the shortest route. I could see how this happened in my particular instance. In my case if you looked at the map and drew a straight line from the origin to destination, the roads that it picked for the shortest route went pretty much right down the line. However the roads were pretty curvy, sort of like a section of the Mississippi river, in general going in a straight line but curving back and fourth. So the calculation for shortest distance probably picks the initial roads by how far it goes off the straight line path, then it calculates the distance and eta from all the actual road data. And since all those curves add to the milage the end result is that it picked the wrong route to be the shortest route. So in the case I saw the fastest route ended up being about 9 miles shorter than the shortest route (and about 2 hours faster).

[doodah]

What's still puzzling me the most, though, is why the routing algorithm is different than the one used in Mapsource or the GPS V? Why change a good thing, only to have it come out worse. I've got to figure that Garmin spent an incredible amount of time developing their routing algorithms and that they would value them as a competitive distinction. What would motivate them to change something that already works well?

I have noticed that the iQue does not offer the ability to select the level of routing quality as they do with their other auto-routing products. This option puts the onus of choosing between quality and speed on the user. Perhaps they've decided to go for speed and just implemented the lower quality version.

Don't get me wrong. I still feel that the iQue does a very good job. It's just that, unlike my GPS V, I generally feel that i need to look twice before I start following a new route.

[apersson850]

The basic platform for the Garmin autorouting functionality is the same between different units. But then they've had to do some tweaking to adjust for different units' characteristics.
They are also considering the ability to make the routing in the iQue more adjustable to the user, by expanding on the adjustments that are offered today.

Besides, from release 3... what the heck, read above! :D