Is the IQ routing effective?

topgazza · Posted: Sun Dec 14, 2008 6:22 pm Post subject:

I agree with your last comment. Its a very clever idea but one that, based on an outdated database, will always be inaccurate..however hard they try and update the database. There are just too many variables to make it anything you could rely on.

The alternative of using it at a very basic, assumption based..."its 8.30am therefore it will be rush hour add 30 mins", level would be a waste of time. We can wrok that out for ourselves.

For me its a gimmick..a "lets get one up on the opposition" marketing ploy...I would prefer a better live traffic service and improved and more regular map updates.
_________________
TomTom 720

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

alanji · Posted: Mon Dec 15, 2008 7:36 pm Post subject:

MrT · Posted: Tue Dec 16, 2008 10:39 am Post subject:

For the example of the M25 above that level of detail would not be required. All that would be needed is the average time between junctions for key times of day. All the matters is the time to the next junction as there are no options or decisions to the route between junctions.
_________________
Drivelux

alanji · Posted: Wed Dec 17, 2008 12:22 pm Post subject:

topgazza · Posted: Wed Dec 17, 2008 1:23 pm Post subject:

..and then things like varying school holidays, bank holidays, festivals.

Just too mnay things to consdier to make it sufficently valuable IMO. As before, better to beef up the on the day live traffic info.
_________________
TomTom 720

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

MrT · Posted: Wed Dec 17, 2008 1:40 pm Post subject:

You could go to a great amount of data in assessing roads but you do not have to go far to get more accurate than before IQ routes. Before IQ routes it appeared that every road could be driven down at a certain speed at any time depending on the type of road.

The first simple addition would be does this road get busy in rush hour and would require one bit of data to flag this. The speed could then be reduced by 50% on roads with this flag during a period called rush hour. Not accurate at all, but far better than before.

Add another bit of data and you can make another general assumption about traffic flows and so on. You would not have to go far before you could have a large range of types of road by speed and time and you would just have to decide which of these categories each road fitted into.

Even if you could build a database measuring every metre of road for every second of the day, I doubt it would be significantly more accurate in predicting journey times than defining 100 models of road type/speed and allocating each road to a type.

Again there is little point in having a really detailed model of traffic flow as it can only take one slow moving vehicle, a badly parked vehicle or a change in weather to totally change the traffic flows in that area.

Going back 20 years ago there was a route planning software called Autoroute that had all the major routes in the UK, roads could also be classified as normal, slow, urban and a combination of these and you could weight your speed by a percentage of these factors and also change road designations yourself. All this fitted on a single floppy disk.
_________________
Drivelux

topgazza · Posted: Wed Dec 17, 2008 1:50 pm Post subject:

Autoroute still exists.

I remember those days as well. It was a coarse guide to how busy a road was and is about as good as it can be now because of all the variables.
_________________
TomTom 720

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

mattcu · Occasional Visitor Joined: Apr 11, 2008 Posts: 25

I've come to the conclusion that it doesn't really matter how detailed the database for IQ routes, they'll never be accurately reflective of reality. There are just too many variables.

As has been mentioned before, some roads get clogged up with school run traffic. But outside of term, these roads can be totally different. I can't see how IQ routes would ever be flexible enough to cope with in-term and out-of-term differences. After all, the route costs only appear to be updated whenever a new map is released.

Another one -- seasonal variations. I work in the middle of the countryside, and we can either get here via the back roads or the main road (but which takes longer in distance and - usually - time). The back roads that my colleagues and I barrel down at 60mph in the summer -- dry and in strong sunlight -- are totally different now. Icy, unlit and with early sunsets, we probably average 25-30mph along the same roads.

So a summers worth of IQ routing data could then be implemented on a new map release at a time when it no longer reflects the route cost of those roads.

Another flaw, at present, is the fact that they cover M-ways and major trunk roads only, AFAIK. Like with the M25 example above -- it doesn't matter how accurate or detailed the data is, if it results in pushing traffic off the 'slow' M25 onto adjacent minor roads which are equally slow but for which no IQ data is stored simply defeats the object.

I'm now of the opinion that TT's time and investment would be better spent on improving the HD Traffic solution to a point where it can naturally succeed IQ routing, at least on a 'live' basis. And on making HD available to a broader base of users via the HD Travel Accessory... ;-)

IanS100 · Posted: Wed Dec 17, 2008 5:16 pm Post subject:

As the amount of data grows it could be updated on weekly or monthly basis via Home, just as speed cameras etc are now. This would all but negate the need to program in seasonal variations

alanji · Posted: Wed Dec 17, 2008 6:43 pm Post subject:

	PocketGPS
	Web