About us

Scan Windows 7

The way forward
Alongside portable music players, GPS has become the second must have consumer technology item in recent years. So what does it take to produce one of these units? We talk to Team Warrior, makers of the AA Navigator range to find out.

What becomes evident is the number of compromises that has to be made both internally and externally. Uday Phadke, Chairman of Team Warrior explains that customers are moving towards personal navigation devices (PNAs) rather than PDAs which have GPS functionality built on top.

This basic need has lead to companies like Team Warrior choosing a processing platform such as Pocket PC, Palm OS, WinCE or Windows Mobile 5 as the basis for their devices. Other bespoke platforms can be used to decrease licensing costs but Pocket PC and its derivatives clearly lead the market with somewhere in the region of 90% market share.

The choice of processing platform becomes a complex decision which has ramifications far greater than just paying a licensing fee to Microsoft or Palm. Phadke riles three trade-offs as if he just made a similar decision five minutes ago. The three vectors; performance, stability and ubiquity clearly have a large impact on the overall user experience. From Team Warrior's point of view performance comes in the form of speed of input/output operations such as reading map data and the availability of connectivity options such as WiFi or Bluetooth.

A cocktail of hardware and software capabilities determines the performance of GPS. Ever found it hard to get a lock on your position before you walk out of the door? That's because of the chip used inside the GPS device. The latest generation of the Sirf chipset called Sirf III will allow you to get signals within buildings and car parks.

Regular users of GPS will know how annoying it is when even in uncluttered surroundings it seems to take ages to get a signal. This is only partly down to the chip within the unit, with the GPS algorithm making a great difference. Even within offices we were able to get a workable signal on one of Team Warrior's units within a matter of seconds whereas my TomTom unit, albeit somewhat older took well over two minutes. Clearly demand of GPS devices has lead to the technology being refined at a great pace.

The GPS algorithm also has a bearing on what sort of device you can use it on. Phadke proudly states that their algorithm will work even on a Samsung chip running at 266MHz, as opposed to others requiring a minimum of 400MHz. This means that a lower specification device can be used thus lowering the cost of entry for users and presumably increasing the potential market for manufacturers.

The actual routing algorithm used to formulate the ideal route needs to efficiently sort through all the map data and categorize roads. Each road has five levels of data with some routing algorithms only the first three. Again compromise is the name of the game, with the quality of route having to be sacrificed in order to lower calculation times.

Regardless of which manufacturer you buy your GPS unit from, the map data or "geo data", is generally acquired from either one of two sources. Both Navteq and TeleAtlas have their merits and while the user rarely gets to make a choice between the two, for manufacturers like Team Warrior the decision boils down to the how detailed each company's geo data is and licensing costs. Once the decision is made its simply like buying a computer game with the geo data arriving on a CD.

A considerable amount of effort goes into differentiating GPS devices and one way to do that is to provide location and non-location aware services. Location aware services are things like speed camera locations, traffic information, location of car parks or restaurants. Non location aware services can be walking maps. Phadke mentions a bespoke solution where the firm could track where their engineers where should they not be able to respond to a signal. Failure to respond would entail that engineer's life was in danger - so less of the George Orwell and more a case of the loving matron.

Differentiation occurs with the user interface, something Phadke classes as "vital". Clearly as GPS is used when on the move the interaction required to operate it should be minimal and shouldn't detract the user away from the job in hand. For that reason certain parts of the data need to be hidden from the user as it would only confuse them, the question is how much do you hide away? Phadke mentions an "easy mode" where only the basic information is available. In the future we should see further customizable interfaces allowing more personality to be presented in what is a currently a rather utilitarian device.

The future of standalone GPS units is somewhat murky. With the majority being used in cars, as more cars come with built in GPS systems clearly a time will approach when supply in an already crowded marketplace outstretches demand. It was clear from our meeting with Phadke and his colleagues that this is a problem but one that has yet to raise its head from the sand.

So to build a GPS system you need to have your sensible hat on. At every stage there requires a balancing trick to be performed and if you have the skills of a trapeze artist you may just be able to conjure up a GPS system which works efficiently.