CoPilot 5 ETA - Tweaking Road Speeds

[lbendlin]

Cool. Now how should the proggie work? A simple form that allows you to specify the speeds and to delete the gpstrip.dat file?
_________________
Lutz

Report Map Errors here:
TomTom/TeleAtlas NAVTEQ

[HielandLad]

Lynnk,

Could you post a printout of the changes you have made? It would give those of us who are less expereinced in these matters an idea of what to do.

Thanks,

[HielandLad]

It would be good if it would also extract the existing speed data so that you know what you are changing it from.

[lynnk]

Ok ... for what it's worth, here is what my USERTRIP.TRP file currently looks like. The only important line is the DefRoadSpeeds line. Below it is the conversion from Hex to Decimal. Note that I am not using the metric units ... I haven't tried switching to see what they look like if you use metric. These numbers have seemed to generate fairly close ETS's for the few trips that I have tested it on. I'm sure I will readjust some after getting some more results. My big problem was that the GPSTRIP file was saving old results that wouldn't let the new numbers work. If you delete the GPSTRIP file, it seems to regenerate it when you create the next trip and store the new numbers ... it is probably ok to leave there until you decide to change the numbers again.


(BTW: Microsoft EXCEL has the functions to allow you to convert between HEX and DEC easily ... use the HEX2DEC and DEC2HEX function. I will try to post in the Download section as Copilot Speed Hex Convert)
=============================
Data Version=5.0.1.3
Start Trip=Default Trip Options
Memo=
DefaultTrip=
HasRoadSpeeds=0
HasStopOptions=1
RouteType=Shortest
VehicleType=Auto
RoundTrip=0
HubRouting=0
SeqType=0
FavorAvoid=1
DistUnits=0
BreakTime=0
BreakWaitTime=0
BorderWaitTime=0
RestrTrucks=1
TranslateAlias=0
StateOrder=1
FerryMiles=1
LoadedCost=11000
EmptyCost=11000
FirstStop=1
HazType=0
BordersOpen=1
LondonZoneOpen=0
TollAvoid=0
NationalNetwork=0
53FootRouting=0
CountriesAsStates=0
ScenicRoute=0
TollAdjust=135
UrbanAdjust=100
ScenicAdjust=100
TollType=Tolls Off
TollDiscounts=0
Length=0
Width=0
Height=0
Weight=0
RoadWeightAdjust=646464646464646464
DefRoadSpeeds=0141413C3705281E14013C373228051E190A


End Trip

End Stop
========================================
MPH
HEX DEC ITEM
01 1 Nothing
41 65 Freeway Rural
41 65 Freeway Rural (no ramps)
2D 45 Divided Rural
1E 30 Primary rural
05 5 Ferries Rural
19 25 Seconday Rural
1E 30 Ramp Rural
12 18 Local Rural
01 1 Nothing
37 55 Freeway Urban
37 55 Freeway Urban (no ramps)
1E 30 Divided Urban
14 20 Primary Urban
05 5 Ferries Urban
0F 15 Seconday (Urban
19 25 Ramp Urban
0A 10 Local Urban

[Bigwig]

Made these changes over a month ago now and have been happy with them. They seem to give fairly accurate estimated trip times for average traffic levels on UK roads. The only problem I have found is that there is only one speed for A Roads regardless of whether they are single or dual.

I found I only had to change the usertrip.dat and deftrip.dat files on my Pocket PC. I just dragged the files to my desktop from the Active Sync Explore Device. Then edited them in Notepad and dragged them back after renaming the original just in case. Changing these values can mean Copilot will choose different routes for the quickest option (but they should be better routes!)

Anyway here they are as hex values:
014646372D05231E19013C3C23190514190F

And in real (decimal) mph:

** 70 70 55 45 ** 35 30 25 ** 60 60 35 25 ** 20 25 15

I didn't change the ferry values.

This is the key I worked on based on what has been posted before and playing with some of the map colours to see which types of roads changed.

nothing
rural motorway
rural motorway (no ramps)
rural divided highway / A roads
rural primary highway / B roads
ferries
rural secondary highway (brown on map)
rural ramp
rural local
nothing
urban motorway
urban motorway (no ramps)
urban A roads
urban B roads
ferries
urban secondary road (brown on map)
urban ramp
urban local

If you don't like my values the Windows Calculator in Scientific view can easily convert HEX to DECIMAL so you can make you own changes.

Have fun.

[PONDEROUS]

Lutz

Dave Burrows has referred me to this thread in connection with my desire to get Copilot 5 to produce routes that are, literally, the shortest; i.e. regardless of the road type.

You, working with a forum member, have produced a method of altering the .dat files in Copilot to tweak the road SPEEDS. Thank you for that: what you say is clear. It has made it easy to apply the same principle and alter the road TYPE preferences used by Copilot in its route calculations when the "shortest" route option is selected. As with the road speeds, I have been able to transfer the files to the PDA and see the difference take effect.

However, I am unable to find any combination of settings will produce a journey that are consistently shortest.

Also, while playing around with Copilot's AVOID or FAVOUR settings does alter the types of roads travelled, at least to some extent, it does not seem to produce a consistent type of journey that matches each combination of the settings. Otherwise, it might be useful to have a few versions of these .dat files on the PDA and to bring them into use for particular circumstances .

I had already concluded from earlier experiments with the desktop version, altering the same values via the User Interface, that it is not possible to resolve the issues that I have described. I am hoping that I am wrong and that you will be able to put me right.

Can you throw any light upon this please?

[PONDEROUS]

A typo in my most recent post might alter the meaning more than some other mistakes, so may I correct it as follows I said, regarding the road type bias settings in Copilot 5:

[lbendlin]

technically speaking shortest routes are guaranteed if all road types have the same speed setting. Have you tried that?
_________________
Lutz

Report Map Errors here:
TomTom/TeleAtlas NAVTEQ

[PONDEROUS]

Thank you for the reply. Yes. I have indeed tried the following:

Setting the bias for every road type to what I would call neutral; i.e. at dead centre between AVOID and FAVOUR. This is in fact where the biases are set when CP 5 for the PDA comes "from the factory". Unfortunately, these settings generate the bias toward main roads that I and other forum members have described in the thread on Copilot 5 UK Routing - hence the nature and sheer number of the posts on that thread.

Setting the biases all to maximum FAVOUR, just in case the logic was that it would then favour every type of road equally. At least on the few test routes that I have put in, the results are no different from having all biases set at neutral.

Setting the biases all to maximum AVOID. This seems to me the most curious of all the situations. For it should give no route calculation at all - since we are supposed to be avoiding every type of road. However, again on my few test routes, the results are the same as with all biases at neutral.

Various combinations of AVOID and FAVOUR. As I said in the earlier post, there appears to be no combination that achieves the neutral situation that I (and, I think, you) would have expected to be available - somewhere.

It is worth noting that the AVOID and FAVOUR settings are on a sliding scales either side of supposed neutral. This means that we can ask CP 5 to, for example, 30% avoid motorways, 20% favour dual carriageways, 40% avoid secondary roads and 50% favour local streets. This sort of adjustment is presumably not intended for tailoring individual journeys. Waypoints, do that on their own, assuming that they work, and in any case these adjustments are not normally available on the PDA. So, just what is the logic here; how are we supposed to know what sort of route we will get the next time we use CP 5 after entering such settings?

All settings of this kind surely have be measured from, or compared with, some kind of base line. Otherwise, it seems to be anybody's guess what any combination of settings is supposed to produce. Does CP5 actually have identifiable baseline settings that will produce shortest routes?

As I have said, I truly hope that I am wrong in my continuing belief that CP 5 is fundamentally flawed, and that I can resolve this and get on with navigating.

I wonder, therefore, if you Lutz, your colleagues or any forum member do have the necessary answers.

I should perhaps mention that ALK are already familiar with what I have written above, and that they have not so far proffered any advice.

[PONDEROUS]

Lutz wrote

[PONDEROUS]

Sorry to multiply the number of posts, but I think this is an essential point.

It would have been nice had equalising all of the road speeds caused CP 5 to generate truly shortest routes. However, isn't "shortest" about distance, and therefore isn't speed an inappropriate factor in the calculation of shortest routes? Also, wouldn't it foul up the "quickest" routes if we were to equalise the road speeds.

[lbendlin]

By setting all speeds the same you essentially eliminate speed from the equation which only leaves distance, and turn penalties (as you know CoPilot has a RV mode that tries to avoid sharp turns). I would recommend you do some more route calculations to verify that those are indeed the shortest routes.

Obviously those settings won't help you with the "quickest" calculations - there you do need the road speeds much more.
_________________
Lutz

Report Map Errors here:
TomTom/TeleAtlas NAVTEQ

[PONDEROUS]

Thanks for that further response.

I think that there must be more to it than this - if there is in fact a solution at all. I have already done, more or less, what you are suggesting and I don't think there is a combination of these settings that will consistently produce shortest routes. I think that ALK would otherwise have responded with a solution, if not on an internet forum then at least to an individual user.

From what I have observed, it is a case of setting a heavy bias towards minor roads when in rural areas, which does at least find shortest routes as long as minor roads only are involved.

For longer journeys that have a number of different types of road available, it seems that there is nothing for it but to adjust with waypoints.

In summary, unless someone does know something other than how to adjust these road type biases, then I think that Copilot users are stuck with having to check every journey until a new version is brought out.

Thank you, nevertheless, for the information on this thread. I now know how to adjust the .dat files and a couple of different versions onn the PDA might prove useful.

[PONDEROUS]

A couple more thoughts, then I'll try to leave you in peace.

The point for me is that while, of course, there will be differences between RV and Auto modes, shortest and quickest are entirely different concepts. Copilot should be calculating shortest distances with the road type biases set at an identifiable, neutral baseline, regardless of road speed settings. Otherwise, any adjustment to shortest routing would inevitably upset quickest routing in the process.

If road speed and type adjustments really are tied together in the way that it appears, then it seems to me that the underlying logic is fundamentally flawed.

[topgazza]

WOW!

Some deep stuff here. Interesting that some clever bods have found a way to improve CP5s apparent lack of what "distance" and "time" means. As in shortest and quickest.

If folk here can do this why can't ALK? I still post as a pretty satisfied user but understand what ol' ponderous is driving at (pun intended)

Its still pretty poor that ALK seem oblivious to the problems. A short "we're working on it" statement would suffice for some. I doubt if we will get it.
_________________
TomTom 720

Nokia Lumia 800 with Nokia Maps, iPhone 4S with Apple Maps (sigh)

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!