The Z-Car Cockpit project
Hardware 
Acquiring data
That's the domain where my knowledge makes default : the electronics. Because I'm a software engineer only. Thus I had to learn from scratch everything, to succeed at the end, obtaining a signal conversion card for input signals comming from the car mading them exploitable by the I/O cards wich I've found on the market.
Sensors
The sensors are essentially those comming with the car, except for speed, considering that it was mechanically transmitted to the original dashboard via a rotary cable, I had to create a home-made sensor reusing an old tachyometer and an ILS relay, generating impulses allowing to determinate distance and speed.
Power supply
On Internet you can find all kind of power supplies already designed for car embedded PCs. But I choose development confort for prototyping : I kept the 220V original power supply and I added a 12-220V power converter. Thus when I develop in bench (at home), I can connect it directly to the common sector (here in France we have 220V-50Hz at home) ; and when I need to test on target (in the car) it is directly connected to 12V car power. I've integrated the 12-220V power converter drectly into the PC's tower where I had much spare space. 
The PC 
I needed something cheap and handfull to be able to crarry out much tests as well in bench while developing that in target to test or show it. The only constraint I asserted myself was that the PC could enter in the trunk of my car, so it will not be visible from outside. I thus chose a standard PC with a micro-ATX motherboard and a mini-tower (not a shuttle).
You shouldn't forget that it is about prototyping and consider that I didn't done yet a reduction for place and cost study for industrialization. I remaind you that I'm working on my onw expenses...

Here is the PC, once installed in the trunk (don't forget that it is a prototype !) : 

click to enlarge   cliquez to enlarge   click to enlarge
Which hard drive ?
Many of you ask me about hard drive and vibrations: during my first tests I used a standard hard drive, put in standby mode by the OS as soon as possible to avoid vibrations problems. But that was not enough, if the weather were too hot or on large bumps, I fell on critical hardware errors blocking the system and then all the dashobard. It was very awkward and that made it unusable.
Thus I finaly choose to use a CompactFlash to IDE converter. This little mervellous thing, that you can find on Internet everywhere, allows you to use a CompactFlash memory card as a standard IDE drive. After some light OS modifications and optimizations of Windows XP, I installed it and sice it could be hot, bumping, cold, etc... everything goes well and remains 100% stable (until now).
Displays
I had a lot of luck for this point of view : in my car dashboard there is very exactly the place for two LCD 7" 16/9 displays set coast at coast. It is thus what I did: the screens that I found have a resolution of 800x480, which makes a total resolution of 1600x480 on the whole dashboard display zone.

click to enlarge   click to enlarge   click to enlarge
Video card ?
I had to inevitably find a video card which is able to manage OpenGL calls with acceptable performances. And of course, to be able to manage two screens it has two video outputs. However it is not the last top-of-the-range video card available in the market, a one from 1 year and a half or 2 years is enough to cover what it must do.  
History

October-November 2006
Embedded PC development: connectors to the car sensors, signal conversion, upstream filtering, etc...

Summer 2006
I devote my summer holidays to develop a skinnable software allowing to compute and display the esential data in a dashboard.

October 2005
The idea was born. While I rolled on a right and tedious road.