As there has been discussion about my navigation project also in other threads, I thought that it is most informative to gather a small Q&A list to get most of the questions answered.

Q: Is the navigation project still going?
A: Yes, it's still alive. It has been more or less work-in-progress since March 2000 when I had the first empeg vs. GPS communication running on my Mk1.

However, for the past half a year, not much has happened since I've been extremely busy with my primary work. Fortunately, things are getting balanced and I hope to spend more time with the navigation soon.


Q: Could you enlighten us on the current state of navigation software?
A: Essentially it's able to project the car's position to the vector map and show the current street name along with street numbers in realtime.

It has a preliminary route planning and guidance feature which lets you to type the destination street name either by using the remote or the rotary knob which after it calculates the shortest/fastest route and starts the guidance. The route planning takes turning restrictions and one-way streets into account.

It does not utilize any PDA device thru IrDA anymore as it turned up to be not so practical solution for in-car use. Instead, all feedback to the driver is given with speech prompts.

Big part of the design was to smoothly integrate the navigation with the existing player software. All speech prompts are mixed with the regular player audio output and therefore the music is faded down to 50% of its normal volume while the navigation speaks the guidance and other information on top.

The display shows the current street name, information about the route such as estimated arrival time/distance left to the destination. It can also show extended GPS information such as speed, number of visible satellites, etc. The navigation software is controlled with the front panel buttons and remote with certain keys fixed to change the focus between the player and navigation software to allow better control. While the navigation software has focus, it has internal menus where various things can be controlled -- meanwhile the player continues to play music on background and remains unaccessible until focus is switched back.

As noticed, there's no web site for the project as I don't see it too useful to hype something that is not publicly available yet (excluding the reference on geek site).


Q: Will it ever be available to general public?
A: That mostly depends if the project gets into a state where it's useful for other people. To make it work with other environments, it will require things like proper installation system (as it requires kernel modifications, relies on having a custom scratch mount point, etc).

The biggest obstacle however being the maps and the interface between the program and various map formats (and each geographical area needing a specific map projection code). Once the map interface works better and has support for other map types, it will be much closer to a state where other people could potentially use it.


Q: Do you need help?
A: At the moment, I don't feel comfortable to turn the project into a massive team effort as I know that it will easily get out of hands without a proper coordination especially since the programming work is not the key fact slowing down the development. It would be different case if I would be starting the project now, but since it's been going on for the past 1,5 years there's a lot of ideas regarding the design/implementation and experiments that I've made which affect to the overall direction I see it going.

However, if at some point I feel like dropping the project, I don't see it as a problem to contribute the existing work to the community. On the other hand, once it's actually possible to use the software in other environment, I'm sure there are many aspects where testing in different countries and geographical areas would help.


Q: Will the source be available?
A: There's no definite yes or no. However, I see it more likely to have a binary version available sooner than the source. Future will indicate where the project goes and which option seems better.


Q: What map types are possible to integrate with it?
A: It's solely based on vector maps which represent the street network using some known map projection formula. However, to get it work with multiple data sets, it seems pretty essential to have the source data always in same format.

For now, it looks like the TeleAtlas maps will be the ones that I'll investigate further. Partially, because it is the only company producing street-level vector maps for Finland. They also have pretty good coverage on Europe and US.

Kim