| View previous topic :: View next topic |
| Author |
Message |
mhb
Regular Visitor
Joined: Dec 21, 2005
Posts: 155
|
 Posted: Wed Mar 08, 2006 8:20 pm Post subject: |
 |
|
It seems to me that we're back on the same tack as the 'inconsistent routing' threads.
On the face of it the f/w algorithm does seem at fault but it's impossible to know for certain unless Garmin openly publish the algorithm used. |
|
| Back to top |
|
 |
Lester_Burnham
Frequent Visitor
Joined: Oct 17, 2005
Posts: 618
|
| Posted: Wed Mar 08, 2006 10:02 pm Post subject: |
|
|
| mhb wrote: | It seems to me that we're back on the same tack as the 'inconsistent routing' threads.
On the face of it the f/w algorithm does seem at fault but it's impossible to know for certain unless Garmin openly publish the algorithm used. |
1. Garmin are NEVER going to do that.
2. Given what's the stated functionality, and the feedback by Garmin to at least two people in this thread, it's clearly apparent, that the fastest route, should be at least the i3's perception of the fastest route - ie not trumped by the shorter option. They've (Garmin) stated to more than one person, here, the criteria that is used for the i3 calculating the fastest route - so quite clearly there's some bug or glitch there - apparent also, because of their inability to explain it.
As you point out, I think it does all come back to inconsistent routing - which some have been attempting to sweep under the carpet, or be in denial about.
I mean if you think about it, it's got two main categories for route calculation - distance or time, so it's clear what the fundamental factors for each selection are - two simple metrics - either distance, or elapsed time (given some forumlaic assumptions). And if selecting the shortest for route selection, actually provides a route which the i3 itself deems as quicker than when fastest is selected, then something is amiss. |
|
| Back to top |
|
|
ride4smilesjonny
Regular Visitor
Joined: Dec 29, 2005
Posts: 222
Location: teesside
|
| Posted: Wed Mar 08, 2006 11:17 pm Post subject: |
|
|
| Lester_Burnham wrote: | As you point out, I think it does all come back to inconsistent routing - which some have been attempting to sweep under the carpet, or be in denial about.
|
_________________
**********************************
Hi ho - Hi ho its off to Arran i go,
with my i3 and a cup of tea hi ho, hi ho
hi ho hi ho.....
**********************************
www.ride4smiles.co.uk
£3500 raised in two weeks! :-) |
|
| Back to top |
|
|
Motty
Occasional Visitor
Joined: Feb 28, 2006
Posts: 11
Location: Cheltenham, UK
|
| Posted: Thu Mar 09, 2006 10:30 am Post subject: |
|
|
Hi everyone, another i3 newby here. I found the same thing at the weekend while testing my i3. At a roundabout near my house, I ignored my i3 and went the route I know is quickest. At that point the ETA decreased by 2 minutes!
My thoughts on this is that it must use quite basic algorithms to calculate the fastest route at the start of the journey. Obviously it doesn't evaluate ALL possible routes you can take. It must just use a set of equations it solves mathematically. So then is the problem as simple as the unit not doing enough iterations to solve the equations? It is a balance between the time the unit takes to calculate a route against how close it gets to the fastest route. This would explain the inconsistent routing problem in the other thread. A slightly different start position could mean it converges to the answer in a different direction. What does everyone else think? |
|
| Back to top |
|
|
DeeDubleYuh
Occasional Visitor
Joined: Feb 03, 2006
Posts: 18
|
| Posted: Thu Mar 09, 2006 7:53 pm Post subject: |
|
|
Well said, Motty. That's exactly what I was about to post... only you explained more clearly than I could have.
The route searching algorithm isn't exhaustive; it tries a few "obvious" routes then decides which is the best (e.g. quickest) among the ones it has found. In searching for the quickest route maybe the algorithm has a rule-of-thumb that says to try the fastest roads first. When it's tried a few routes (giving preference to fast roads) it stops looking for other options and shows you the quickest that it has found.
It's not very encouraging that members have found that Tom Tom software has a better algorithm. I've read in many places that people prefer Tom Tom software to anything else.  |
|
| Back to top |
|
|
noddy
Occasional Visitor
Joined: Jul 07, 2004
Posts: 17
|
| Posted: Thu Mar 09, 2006 7:59 pm Post subject: |
|
|
Hi Motty
I think you might have hit the nail on the head. The calculation system may be a lot more basic than more expensive units and perhaps its not evaluating all the possibilities and therefore for some unknwn reason throws a wierd one in from time to time. Someone at Garmin must know the answer but perhaps they choose to play dumb on the subject for marketing reasons. It still doesnt explain why mine wants to send me the longest route and slowest to this one particular destination. Yet when I put another destination in just a couple of miles further then it sends me the right way even when the route travelled will be identical apart from the extra 2 miles!! Theres no logic to that at all. Its got nothing to do with the positon of satellites or where my car is parked because I'm doing this inside the house without a signal. |
|
| Back to top |
|
|
zogman
Frequent Visitor
Joined: Sep 05, 2005
Posts: 1417
Location: swindon uk
|
| Posted: Thu Mar 09, 2006 8:22 pm Post subject: |
|
|
noddy , you say that you do this inside the house with no signal...does this mean that you are in simulation mode..if thats the case maybe simulation mode gives different routes and times than 'real' life ....
_________________
*************************** |
|
| Back to top |
|
|
Motty
Occasional Visitor
Joined: Feb 28, 2006
Posts: 11
Location: Cheltenham, UK
|
| Posted: Thu Mar 09, 2006 8:29 pm Post subject: |
|
|
| DeeDubleYuh wrote: |
It's not very encouraging that members have found that Tom Tom software has a better algorithm. I've read in many places that people prefer Tom Tom software to anything else. |
I think it is slightly unfair to compare the i3 to Tom Tom because the i3 is a cheap basic model made from cheap basic components. I'm guessing its processor will be a lot less powerful than the one in a Tom Tom and so the i3's routing algorithms must be simpler |
|
| Back to top |
|
|
noddy
Occasional Visitor
Joined: Jul 07, 2004
Posts: 17
|
| Posted: Thu Mar 09, 2006 9:25 pm Post subject: |
|
|
| zogman wrote: | | noddy , you say that you do this inside the house with no signal...does this mean that you are in simulation mode..if thats the case maybe simulation mode gives different routes and times than 'real' life .... |
Its the same out on the road which is how it came to light in the 1st place. I was driving home and it wanted to put me onto the M69 instead of taking a relatively direct A road. Its that detour that would put an extra 4 miles on the route. Which is why I came home and started messing about with it to find out why it wanted to do that.
Ive tried putting in different destinations all northside of Leicester (i.e. taking a virtually identical route) and it will take me there the sensible route but as soon as I put Coalville in as the destination it wants to take me 4 miles out to the west on the M69 to eventually join up with the M1 at the place I would normally join it for all other North Leicester destinations |
|
| Back to top |
|
|
swing
Pocket GPS Verifier
Joined: Nov 04, 2003
Posts: 2225
Location: Bedfordshire, UK
|
| Posted: Thu Mar 09, 2006 10:06 pm Post subject: |
|
|
| noddy wrote: | | The calculation system may be a lot more basic than more expensive units and perhaps its not evaluating all the possibilities |
To do the calculations required takes both processing power, time, and memory, all of which I suspect are limited on the i3, hence why it doesn't always offer the best route. Strangely, I don't see people complaining about the routes on the c, 2xxx or Nuvi ranges, so it would appear to be limited to the i range, which if the issue is as believed might make sense for the 'budget' range. |
|
| Back to top |
|
|
Motty
Occasional Visitor
Joined: Feb 28, 2006
Posts: 11
Location: Cheltenham, UK
|
| Posted: Thu Mar 09, 2006 10:38 pm Post subject: |
|
|
And without knowing the algorithm Garmin use, I think it is going to be very hard to solve this problem.
Maybe it splits the route up into a few major sections and calculates from route of each section independantly and for your problem the major sections it chooses is poor?
Who knows? ...Only Smarties have the answer...or should that be Garmin... |
|
| Back to top |
|
|
Lester_Burnham
Frequent Visitor
Joined: Oct 17, 2005
Posts: 618
|
| Posted: Thu Mar 09, 2006 10:38 pm Post subject: |
|
|
| Motty wrote: | | DeeDubleYuh wrote: |
It's not very encouraging that members have found that Tom Tom software has a better algorithm. I've read in many places that people prefer Tom Tom software to anything else. |
I think it is slightly unfair to compare the i3 to Tom Tom because the i3 is a cheap basic model made from cheap basic components. |
And?
It should still be competent at it's core function - which is navigation.
As a budget unit, it may not have the fuller features of more expensive units - faster CPU, bigger screen, touch screen, recharging circuits, bluetooth, TMC input, many more configuration options. I can forego all of those, in realisation that I have bought a budget unit - but it should do it's core function, reliably and consistently. That's not unfair, unreasonable or exceptional expectation.
| Motty wrote: | | I'm guessing its processor will be a lot less powerful than the one in a Tom Tom and so the i3's routing algorithms must be simpler |
That's a non-sequitur. It's processor may not be as quick or as powerful as the more expensive units - that's a distinct possibility. But that doesn't mean that the routing algorithms should be a world apart. They may not have as many configurable / tweakable inputs, but the core principles should be very similar.
And if the i3's processor is slower, it should just mean it takes a little longer to provide routes and recalculate, than more upmarket units. |
|
| Back to top |
|
|
Lester_Burnham
Frequent Visitor
Joined: Oct 17, 2005
Posts: 618
|
| Posted: Thu Mar 09, 2006 10:40 pm Post subject: |
|
|
| noddy wrote: | Hi Motty
I think you might have hit the nail on the head. The calculation system may be a lot more basic than more expensive units and perhaps its not evaluating all the possibilities and therefore for some unknwn reason throws a wierd one in from time to time. Someone at Garmin must know the answer but perhaps they choose to play dumb on the subject for marketing reasons. It still doesnt explain why mine wants to send me the longest route and slowest to this one particular destination. Yet when I put another destination in just a couple of miles further then it sends me the right way even when the route travelled will be identical apart from the extra 2 miles!! Theres no logic to that at all. Its got nothing to do with the positon of satellites or where my car is parked because I'm doing this inside the house without a signal. |
I'm with you on that - there's no logic to it, which is why it's difficult to accept. I've found much the same thing. And it seems you'll either get (effectively) shrugs from Garmin, or apologists saying you should be grateful for what you get for the money, and it's best not to rock the boat too much. |
|
| Back to top |
|
|
Skippy
Pocket GPS Verifier
Joined: 24/06/2003 00:22:12
Posts: 2946
Location: Escaped to the Antipodies! 36.83°S 174.75°E
|
| Posted: Thu Mar 09, 2006 11:03 pm Post subject: |
|
|
| Motty wrote: | | My thoughts on this is that it must use quite basic algorithms to calculate the fastest route at the start of the journey. Obviously it doesn't evaluate ALL possible routes you can take. |
Yeah, I think you have hit it on the head there. I think they have a heirachy of roads that they examine to find a route rather than doing an exhaustive search of all possible routes. If they find a route using the main roads then it will use it and presume that if there is an A road from point a to b then it's probably the fastest route.
In somewhere like London, they can examine 10% of the roads and get a route that is within 5% of the fastest route and that is good enough.
There has to be a point where the unit decides that the route is "good enough" and stops searching and gets on with navigating. The ETA decreasing by 2 minutes proves this point. If the sat nav had tried harder it would have found that route, but there has to be a trade off between a quick calculation and a good route.
_________________
Gone fishing! |
|
| Back to top |
|
|
Motty
Occasional Visitor
Joined: Feb 28, 2006
Posts: 11
Location: Cheltenham, UK
|
| Posted: Thu Mar 09, 2006 11:16 pm Post subject: |
|
|
| Lester_Burnham wrote: |
It should still be competent at it's core function - which is navigation.
|
It is competent....it gets you there!
| Lester_Burnham wrote: |
That's a non-sequitur. It's processor may not be as quick or as powerful as the more expensive units - that's a distinct possibility. But that doesn't mean that the routing algorithms should be a world apart. They may not have as many configurable / tweakable inputs, but the core principles should be very similar.
And if the i3's processor is slower, it should just mean it takes a little longer to provide routes and recalculate, than more upmarket units. |
but the routing algorithms WILL have to be quite different if it uses a budget processor and you want it to calculate a long route in less than 15 minutes! Remember it also has to recalculate the route when you take a different turn to what it suggests. This recalculation has to be quick or you will be 5 turns further along the route, possibly somewhere different to where it wants to take you so has to recalculate the route again. The unit is full of compromises. Calculation time is more important than getting the absolute fastest route. More people would complain if it was slow to calculate a route. |
|
| Back to top |
|
|
|
![]() Posted: Today Post subject: Pocket GPS Advertising |
|
|
We see you’re using an ad-blocker. We’re fine with that and won’t stop you visiting the site.
Have you considered making a donation towards website running costs?. Or you could disable your ad-blocker for this site. We think you’ll find our adverts are not overbearing!
|
|
| Back to top |
|
|
|
Forum FAQ
Search
Usergroups
Profile
Log in for private messages
Log in
