UK Shootout – TomTom2 vs. Destinator 3
by Mick Jennings

 

This document compares two leading road navigation systems for the Pocket PC:

 

- TomTom Navigator 2.24 (TT2)

- Destinator 3 (D3)

 

The hardware used is an Ipaq 2210 equipped with CF 512Mb and SD 512 cards. The GPS used for these tests is the TomTom Bluetooth, although at a later date I intend to repeat the tests using the Fortuna CF Pocket XTrack.

 

Both packages were installed in the default directories on the Ipaq, with UK maps for both packages installed onto the CF card - there is a known problem with the Ipaq BT stack when reading data from SD cards.

 

A soft reset was performed before and after switching packages to perform the tests. After the soft reset the Ipaq had 10 Mb “Storage” free and 24 Mb “Program” free. This was checked and was the same for each program.

 

The Tests

 

In selecting the tests I have tried to emulate real-life requirements from my own experiences. In my opinion there are a number of obvious and very common requirements from a road navigation package which I would summarise as follows.

 

·         Show the current location on the map

·         Find a destination

·         Plan a route

·         Navigate the driver to that route, with clear and unambiguous voice and display guidance

·         Give a reasonable guide to ETA

·         Cope with the driver missing a turn, quickly providing guidance as alternative course

·         Plot alternative courses to avoid roadblocks and traffic conditions

 

All times referred to below are average times - all tests were repeated 10 times. For interest, the variance among the tests results was extremely low - the times recorded were virtually identical with each iteration of the same test.

 

As well as measuring specific test results I “lived with” each product for a week and some of the more subjective feedback comes from this.

 

How the packages performed

 

Location Display

 

Both packages interfaced successfully with the BT GPS and displayed the current location without problems.

 

Find Destination

 

Finding a destination by it's address is significantly faster using D3 than TT2. For example, looking up the address 60, St Mary Axe, London took 29 seconds (including keystrokes) using D3's City-Street method. Looking up the same address in TT2 (which only offers the City-Street method) took 90 seconds. The address lookup in D3 is actually a joy to use. Start typing the name of a city or street and the displayed list is immediately filtered with no noticeable delay. TT2 on the other hand often misses keystrokes because it’s so busy working on your last letter. The majority of the time shown above is spent waiting for the lists to refresh and it is highly frustrating to use.

 

Finding a contact with the same address took 56 seconds under D3 as opposed to 58 seconds under TT2 (which has to be driven from the Contacts application) so no real difference there.

 

Plan and display a route

 

I tested this functionality using four routes:

 

-          long distance and mostly motorway

-          medium distance and complex with several possible alternatives

-          shorter distance across country

-          within a city.

 

All tests are using "Quickest" options and with default settings. These times do not include keystrokes, i.e. they are timed from the final keystroke to the system returning control.

 

Long route :- The route 99 Dewsbury Road, Leeds to 60 St Mary Axe, London took 163 seconds on D3 vs. 235 seconds on TT2.

 

Complex :- The route 1 Bridge Street, Penistone to the junction of Red Lion Street and Mitre Fold, Wolverhampton took 118 seconds on D3 vs. 506 seconds on TT2.

 

Cross-Country :- The route 1 Bridge Street, Penistone to 11 Pinfold Crescent, Tideswell took 60 seconds on D3 vs. 53 seconds on TT2.

 

Short route :- The route 99 Dewsbury Road, Leeds to 48 Albion Street, Leeds took 55 seconds on D3 vs. 11 seconds on TT2.

 

This confirms my overall observations, namely that TT2 is extremely fast at short distances but can take an age to calculate long or complex routes.  D3 however is better at long and complex routes but appallingly slow on short distances. I will explain below why I believe this to be so important.

 

 

Navigation and Guidance

As far as map display is concerned both packages score well in terms of clarity and usability. I favour the "Birds Eye" view in D3 and the 3D view in TT2. Overall I find TT2 to have the easier display to use while driving. D3 displays it's ETA information along with velocity etc on a rolling status bar (blue on the screen shot). If you need to know your ETA you have to look at the screen until the ETA pops up – not recommended while driving.

 

TT2 scores way above D3 here(at least in the UK), for one major reason : it correctly prompts on

 roundabouts. I gather the states doesn’t have these, but they are basically a circular road where the roads join radially. All traffic goes around the same way and you just choose where to get off. They often have multiple lanes on them. Approaching a roundabout, TT2 will give clear guidance such as "Cross the roundabout, 3rd exit, then enter motorway" (see the text on the screen shot). D3 on the other hand treats each possible exit as a road junction, prompting each with “Right Turn”. It is often still showing "Right Turn" when you want the next exit (a “Left Turn”), which is highly misleading. If you were working by voice prompts alone and in an unknown place you would soon go wrong.

 

In my view this is one of D3's two biggest flaws - more later.

 

With D3 every motorway junction is prompted with “Keep Right”. Once on a road TT2 doesn’t prompt you to exit until you need to. At first I couldn’t understand D3’s behaviour here. However, while writing this paper I realise one very significant reason for having all of those junctions in the route. It’s to do with diversions and I’ll discuss it later.

 

ETA

 

Both TT2 and D3 are pessimistic in my experience, although D3 far more so. I regard this as a minor fault however as I always prefer to allow plenty of time (ed - this is particularly a problem with D3 that is systematically 20 to 50% "slow")

 

Recalculation after missed turn

 

This is D3's second and probably most significant major drawback. When you miss a turn with TT2 it realises VERY quickly that you've done so. It seems to work by following your path along the planned turn (shown on screen by a green arrow, see screen shot). If you vary from the green arrow it knows you've gone wrong and immediately looks for an alternative. In doing so it seems to try to get you back to one of the planned points on the pre-planned route, which means that it’s recalculation is virtually instantaneous. In this way it copes very well whether your route is short or long. If you go wrong either at the beginning or end of a long route it will react quickly to get you back on track.

 

D3, however, seems to take forever to realise you've missed the planned route. You have to be some distance away before it recalculates. Worse still, it seems to do a full recalculation of the complete route ! Imagine trying to find a hotel in Bristol and missing the turn. D3 took so long to sort me out that I had to park up while it worked out a route for me. I was only half a mile away as well, and this is why D3's slowness to calculate a short route is so important. Often when you go wrong you are at least in the right city and sometimes just a few streets away. When I found the hotel I deliberately went around the same loop again using TT2. I  went past the same missed street but before I reached the next one (100 yards) TT2 was telling me to take it. That's what you need when you're lost.

The other problem which affects navigation is of course the quality of the map data. TT2 uses the

 Tele-Atlas maps. The worst problems I found were a few dirt-tracks marked as passable and the new Birmingham M6 Toll Road not shown. D3 uses the NavTech maps and I found several more problems with these, usually roads marked passable which now have office blocks built over them (ed - on average, the NavTech maps will be more efficient for driving as they commit to updating the major roads every 6 months)

 

Neither package allows map editing, although they have very different ways of marking routes as impassable. In D3 this is done by entering the road name. This doesn't work where the road is unnamed - which most roads small enough to be built on are. It does not seem possible to mark an unnamed road as impassable in D3. TT2 allows you to draw an "avoid" area on the map which gets over the problem.

 

I should also note that I found the routes which D3 gave me to be much more like the kind of routes I would choose for myself, whereas TT2 seems to stick rigidly to motorways.

 

Diversions

 

D3 has a neat feature whereby you can scroll through to any planned turn on the route and tell it to avoid that road. Very cool.

More on this below.

TT2 has two good options. You can "Avoid" a road which is on the route. However the options presented only include the major roads. The other useful feature is the "roadblock", where you tell TT2 that the road is blocked for, say, 2 miles and it will plot a route around it. Much used on my

 journeys I can tell you !

Both methods have their advantages. I’d love to have the best of both of course. Overall I think I slightly prefer the TT2 method as it's easier to use quickly and can be mapped to hardware buttons. However I'd like to see the option to tell it that the route is blocked some way down the line - the roadblock feature only works from where you are when you use it and you often hear radio announcements telling you of blockages ahead. Imagine the radio telling you that your route is blocked 15 miles ahead. You may well be coming up to a junction which you should take instead. If you need to follow a motorway for 50 miles and part of that road is blocked there’s no way of handling this in TT2.

 

D3, on the other hand, allows you in theory to scan ahead through the list of junctions and pick which one to avoid (see screen shot). Since each motorway junction is listed as a separate “turn” you could work out which one to avoid and if appropriate re-route around the problem. I haven’t tested this functionality, but as D3 does a complete recalculation I wouldn’t be at all surprised to see this work. This explains, for me, the otherwise confusing behaviour of listing each junction as a “turn”, even where it’s telling you to stay on the motorway – it gives you finer control as to which part of the route to avoid !

 

Conclusion

 

I bought D3 because I was very unhappy with the route planning times taken by TT2. However, in my opinion D3 has two fatal flaws which make it an impractical solution:

 

-          the time taken to correct the driver after a missed road (even on a short route in a city)

-          the misleading guidance given on roundabouts

 

TT2 also has it’s flaws. The time taken to calculate a long or complex route is, well, disappointing to say the least – especially as D3 manages to calculate the same route in around one quarter of the time. Bear in mind here that we are talking about 160 miles as a “long distance”. However, this can be overcome by calculating the route in advance (the option to save and load routes would be enormously helpful given this problem). It may well be that TT2 stores information about the route turns which D3 doesn’t and this additional planning time may give TT2 the edge when it comes to recalculation. If this is true then the investment is well worth it, though frustrating if you’ve forgotten to plan the route ahead of time.

 

Reluctantly I'm probably going to go back to TT2. I’ll just have to make sure I plan routes before I set off.

 

Mick Jennings

 

 

 

Update as of 13^th Dec 2003

 

After taking the above screen-shots I did a complete hard reset on my Ipaq and installed TomTom 2.24 with the GPS 2.06 drivers (these are used because 2.07 and later don’t allow the driver to be accessed by any other program and I want to use off-road programs like OziExplorerCE). The Destinator maps were removed from the CF card. I wanted to make sure that the installation of D3 wasn’t affecting the TT2 calculation times.

 

I also made sure I didn’t have any “Avoid” areas – it occurred to me during the writing of this comparison that they may have impacted on the calculation times.

 

Unfortunately it still took 510 seconds to calculate the “complex” route.

 

 

 

Copyright owned and retained by Mick Jennings - mick.jennings@clara.co.uk

publication permission granted to www.GpsPasSion.com in recognition for the excellent service provided to the GPS community.