How to prevent A-road avoidance!?

[PloddinPedro]

An example is the A1141 in Lavenham.  To stay on the A1141 you must turn off the through road.  Again, you can't tell from the map, although if you go to a satellite photo and zoom right in you can see the Give Way lines.

that's very much my point - using Tracks, you don't have any of that hassle - you just draw where you want to go and follow the line!

I use routes for nice big light up at night instructions on how to take junctions.

With Tracks, you just look at the screen, same as daylihgt. A backlight of 10% is easily bright enough, it's always on and the Next Turn screen doesn't obscure the road ahead. Like I say, I still don't see the attraction of Route for Audaxes.

I do however, take FF's point that the "GO TO" routing function will indeed get you out of trouble one way or another, if nothing else is available, and that can in certain circumstances be a boon. But it does depend on accurate mapping and my experience with the local City Navigator mapping, which shows bridleways as paved roads and omits other normal roads altogether, moves me to not bother with it. Part of my attitude I admit, derives from the disappointment I felt when I realised the Routing function was so flawed. When I first got the Garmin I thought it was just brilliant and it wasn't until it had taken me all round the houses on the Ruislip 300k that I started to fall out of love with it!

[Euan Uzami]

It should do but it's a computer

  it should do because it's a computer  
They're not random, they haven't got a mind of their own. They perform calculations based on a set of rules, which are predetermined - in routing, those rules are that there is a matrix of "nodes", each of which is connected to a number of adjacent nodes, where each connection has a 'cost' associated with it, and it routes by walking from a start node to every adjacent node, and from each one of them to each of its adjacent node, and so on... until it gets to the target node.
Obviously it also prunes certain branches, as an optimization, as to calculate the cost of all the permutations would take forever, but essentially I simply do not see how the number of choices for connecting two points which only have a couple or three nodes between them can yield enough similar routes for there not to be one obvious one that stands out. Which is the 'correct' one, i.e. the one that it's obvious to a human has the least cost. In other words, it should be possible to feed points into it that is friendly to the way it works, such that it is obvious to it how to pick the route that we humans consider the correct one.
Certainly the router we have got at work would do that, and I can't imagine garmin would have a worse one that some small backstreet company like us, so it must be something else - I suspect to do with the translation from the lat long of the point you give it to the node/connection that it considers that to be part of.
A translation which, remember, is a non-trivial step since the connections (roads) are as far as the algorithm is concerned, the width of a mathematical line - in other words, infinetessimally thin - so it must assume it to be on the one that is the smallest eucldean distance away. No need for fuzzy logic there, but the only place I can see where it could go wrong if it is calculating the costs between nodes correctly.
It's also possible that with the unit probably running low-power cmos, the pruning is doing something crazy. This is obviously controlled by the 'best/better/quicker/quickest' setting, but i wonder if even on best, it is still a bit arbitrary?

, it won't work properly, they never do. (Moi, cynical?) And anyway if you have to put in so many points, you've basically got a Track!

I agree that the confidence that you gain from not allowing its algorithm to mess you up is an overpowering reason to always use tracks. The only thing you don't get is the "distance to turn" countdown, which is the main reason i like the turn by turn mode.

[andygates]

The problem is that there's limited user tweakability: no scale of highway odiousness and no way of personally changing the 'scoring' that roads and turns get.  The presets are balanced at best guess, but it's as much art as science, and inveterate tweakers like this forum's GPS mob could effortlessly come up with different cycling profiles (eg Vehicular Cyclist - like a car, only use bike routes where they're a distance saver, avoid dual carraigeway; Pub Pootle - avoid busy roads, prefer scenic routes; Commuter - shortest distance wins over all; and so on.  Look for 3D mapping to allow "avoid steep hills", "flattest route", "most challenging" and more in the next generation of devices!

Andy, maybe there is with the OSM maps. I don't know if this will work in practice, but you could maybe 'adjust' routing preferences by changing the road class / road speed for the mkgmap / OSM maps.

Oops, just spotted this.  Yes, in theory you could do that.  Of course you'd end up with a specialised map and the OSM cycle maps may already do this (I don't think they did last time I checked).

[Manotea]

I use routes for nice big light up at night instructions on how to take junctions.

With Tracks, you just look at the screen, same as daylihgt. A backlight of 10% is easily bright enough, it's always on and the Next Turn screen doesn't obscure the road ahead. Like I say, I still don't see the attraction of Route for Audaxes.

I used a 'bare track' on the Upper Thames and had several false turns when I misinterpreted what the track was telling me. Could be that I'm not used to navigating by track. I like routes though. Maybe its a techy geeky thing, but I enjoy using the technology and seeing what it can do. Overtime I've developed a style of coding routes which has proven reasonably reliable and incorporates waypoints with labels (town names mostly) which I can match up with real world signage. Also, using the software I do (Mapsource and WinGDB3), I code the route from which I generate a track, so the route comes first. YPYPATYC. YMMV.

[PloddinPedro]

It should do but it's a computer

it should do because it's a computer  
They're not random, they haven't got a mind of their own. They perform calculations based on a set of rules, which are predetermined .......

Ah yes, but ........ mere ignorant yumans like wot I am can't figure out what those rules are!

My first experience with Mapsource Routes and the GPS unit exemplifies my point. I composed a Route on the Desktop using the Mapsource application I bought with the GPS and downloaded it to the unit and went for a ride. Twenty miles and suddenly I'm going completely the wrong way, on a path which will double the forty mile journey. I telephoned the Support man at Garmin. Him:"ah well sir, you see the unit receives only the waypoints from Mapsource, not the whole Route and since the algorithms in the unit differ from those used in Mapsource it will  come up with a different Route" Me: "so I can never know what Route the unit will generate then?" Him: "no sir. Sorry" And this is a combination of products Garmin manufacture and sell ostensibly to be used with one another!

As I said earlier: Moi, cynical?

Of course, since then I've learnt that it's possible to reduce the chances of this happening by being more thorough with the route building, but for me it does rather detract from the appeal of automatically generated Routes.

[nmcgann]

Good point by PloddinPedro.

I fell foul of the fact that Mapsource and the Gps independently calculate the route pretty much the first time I used it. I'd planned a circular route which looked great in Mapsource, but the gps wanted me to head home almost immediately 

It's ok if you stick lots of waypoints in though.

[Euan Uzami]

I use routes for nice big light up at night instructions on how to take junctions.

With Tracks, you just look at the screen, same as daylihgt. A backlight of 10% is easily bright enough, it's always on and the Next Turn screen doesn't obscure the road ahead. Like I say, I still don't see the attraction of Route for Audaxes.

I used a 'bare track' on the Upper Thames and had several false turns when I misinterpreted what the track was telling me. Could be that I'm not used to navigating by track. I like routes though. Maybe its a techy geeky thing, but I enjoy using the technology and seeing what it can do. Overtime I've developed a style of coding routes which has proven reasonably reliable and incorporates waypoints with labels (town names mostly) which I can match up with real world signage. Also, using the software I do (Mapsource and WinGDB3), I code the route from which I generate a track, so the route comes first. YPYPATYC. YMMV.

my future strategy is going to be to use a track, but only for the purposes of making sure i'm still going the right way, rather than to point me the right way in the first place - that'll be for the routesheet to do.
This should work perfectly for the classic situation where you've got two left turns that start from (almost) the same place, but gradually diverge in different directions. The track might not be able to tell you which one to take at the point of turning, but once you're a bit of the way down it, you'll soon know if you're on the wrong one as you'll be getting further and further away from the line.

[Manotea]

my future strategy is going to be to use a track, but only for the purposes of making sure i'm still going the right way, rather than to point me the right way in the first place - that'll be for the routesheet to do.

Its long occured to me that the best way to enjoy an audax - in fine weather anyway - is to follow the routesheet and use the GPS as an added value bike computer.

This should work perfectly for the classic situation where you've got two left turns that start from (almost) the same place, but gradually diverge in different directions. The track might not be able to tell you which one to take at the point of turning, but once you're a bit of the way down it, you'll soon know if you're on the wrong one as you'll be getting further and further away from the line.

Thats what I was referring to as a false turn. Not a big deal but a tadge annoying when you're riding ahead of a group and then have to play catch up. There is a moral there somewhere.

Something else to watch out for is the gps or rather the rider getting confused when theroute/track visits the a control more than once such as on  a figure8 or there and back route...

[andygates]

I think the confusion here is that you want the thing to go along a nice route, rather than any route that gets you there.

Canonically, the answer must be "use more waypoints when creating your route"

But if you know where you're going, asking something to work out where to go is a bit redundant.  Tracks are more suitable.  Or, if there's a route sheet, just use it as a magic map.

The routing comes into its own when you get lost and have to say "machine, find me a station" or "machine, take me to Ipswich".  Not when you're comfortably on-piste.

[Manotea]

If a Garmin GPS reliably generated the same route as Mapsource I doubt we'd be having this conversation. It simply wouldn't be an issue.

[Lycra Man]

Manotea - I couldn't agree with you more. I know there are many people who have been able to get their GPS machines to work as they wished, but in my case, my Garmin Etrex Legend Cx so often tries to route me the 'wrong' way on training rides, (where I know the terrain), I don't really trust it to guide me along a route I don't know - which was the main reason for buying it in the first place.

If Garmin could get the handheld device to accept and repeat the route beautifully and carefully created on my PC, I for one would be delighted. Until then, I will keep on trying, and following the advice of others.

Lycra Man

[frankly frankie]

The routing comes into its own when you get lost and have to say "machine, find me a station" or
"machine, take me to Ipswich". 

 

As a credit-card tourist in France, the following sequence performed at about 4:30pm is a killer GPS feature -
Find
Hotels
(pick nearest ie 1st one in list)
Go To

Actually the sensible tourist would also avail themselves of the phone number that it gives you - but I don't do mobile phones, more fool me.  So, about 6 weeks ago, we cruised up to the 'nearest' hotel, to find it was a full-blown chateau, with rooms at about 300 euros per night 

[rogerzilla]

The City Navigator NT map also has a stack of false "no through roads" and banned turns which don't exist.  I made about 20 attempts to put the Exmouth Exodus route into it but it kept doing U-turns at one particular crossroads.

We pay for this crap, which is more annoying than if there were an error in OSM.

[andygates]

about 6 weeks ago, we cruised up to the 'nearest' hotel, to find it was a full-blown chateau, with rooms at about 300 euros per night 

Ha!

Going back to "why doesn't Mapsource give me the exact same route?" don't forget that the GPS can be told all sorts of parameters including how much effort to put into the calculation (best to fastest).  That's a lot of variation even before we get to the GPS units being shipped with one set of routing algorithms and more or less set in stone (you'd have to do a device update to change it), while Mapsource could be updating with every update.  So it's not just blithering incompetence on the part of the Garministas in their Ivory Towel.   

[Manotea]

Following Andy's post I had a look round and updated the firmware on my eTrex from V3 to V3.10. Doubt it will any difference, TBH, I'll be pleased if my eTrex still works.

Regardless of that, Andy has a point regarding the number of routing combo's. However it seems reasonable to me that a route generated under Mapsource could/should include the associated Garmin GPS firmware/software routing config details and/or some form of alert to indicate that mapsource and gps routing config options are not aligned.

OK this might only be relevant when downloading routes from Garmin Mapsource to a Garmin GPS but there's nothing wrong with adding a bit of proprietory functionality. IMO.

[Martin]

Audaxing, I don't use routes for 'on the fly' routing, I use routes for nice big light up at night instructions on how to take junctions.

ditto; a 200 takes about 10-20 mins to plot a bomb-proof route that will slavishly follow the route if using Mapsource on your pc, the trick is to plot the viapoints just after each critical turn.

Be aware that it often doesn't alert you to SOXs and a few "R, appears SO" even on big A roads unless there's a slight staggered X so you do also have to keep your eyes open.

[PloddinPedro]

Regardless of that, Andy has a point regarding the number of routing combo's. However it seems reasonable to me that a route generated under Mapsource could/should include the associated Garmin GPS firmware/software routing config details and/or some form of alert to indicate that mapsource and gps routing config options are not aligned.

I most emphatically agree with this. There may be good technical reasons (I assume) why Garmin markets products which don't work together as well as their User Guides suggest. But to leave the buyer in ignorance of this and simply allow them to find it out in the course of ownership I regard as an insult and must surely qualify as lousy marketing in the long term. I for one may or may not consider one day getting a Dakota say, or similar, but I'll be doing so only after the most exhaustive reviews by early adopters and only if no other manufacturer offers the same functionality and I'll be starting from the assumption that the auto-routing function doesn't work, until the forum users assure me otherwise.

[frankly frankie]

And you'd be right about the Dakota - auto-routing doesn't work, I spent a few hours proving that to myself yesterday.  I'll be optimistic and put it down to 'immature product'.

the GPS can be told all sorts of parameters including how much effort to put into the calculation (best to fastest).  That's a lot of variation even before we get to the GPS units being shipped with one set of routing algorithms and more or less set in stone (you'd have to do a device update to change it), while Mapsource could be updating with every update.

Yes, a lot of variables.  But would we want any fewer?  I suspect not.

However it seems reasonable to me that a route generated under Mapsource could/should include the associated Garmin GPS firmware/software routing config details and/or some form of alert to indicate that mapsource and gps routing config options are not aligned.

Good idea.  At present you don't even have matching sets of options between the two - Mapsource needs a 'preferences' option to describe the target GPS.  But I suspect Garmin are losing interest in Mapsource.

I have occasionally seen a prompt on the GPS if the routing was generated on a different map to the one loaded in memory.  I got this when I'd moved to Metroguide v9 on the desktop, but had left the GPS still loaded with v8.

[andygates]

It's the checkbox "pull routing preferences from device" that's needed. 

That's non-trivial if not coded-for already, but certainly doable. 

[frankly frankie]

Or how about a 'learning mode' for the GPS.
Ride your route from A-B and tell the GPS 'thats how I like to do it'.
Do some other route and add to the knowledge base already stored.

[Greenbank]

Or how about a 'learning mode' for the GPS.
Ride your route from A-B and tell the GPS 'thats how I like to do it'.
Do some other route and add to the knowledge base already stored.

Not everyone likes the same kinds of roads.

Not all roads are the same at all times of the day. I avoid some roads at rush hour but would happily ride them outside that. I'd avoid some A-roads during the day but happily ride them at 2am. I'd happily ride some A-Roads during rush hour (when the traffic is slower) than outside that when it's used as a dragstrip by twunts.

etc, etc.

[frankly frankie]

Not everyone likes the same kinds of roads.

I meant as an individual, just me-and-my-GPS sort of thing.
And time of day could be part of the parameters learned.