As well as the imaging devices, all the cars used a combination of inertial detectors (accelerometers) and speedometers to add a further sense of movement to the overall picture. When used in conjunction with GPS tracking these enable the cars more to accurately plot their exact position on the course and ensure that, when passing through tunnels and other GPS blackspots, the cars are always aware of where they are.
As you can see, some competitors used masses of sensors to gain as much information as possible, while others went for a 'less-is-more' approach. However, most of the same type of sensors and other monitoring devices were employed by all the teams and what really differentiated the teams was how they interpreted the data and to see this we need to take a look inside the cars.
It was fascinating to see the different ways the teams had chosen to interpret all the data being sent from the sophisticated sensors. While MIT's Talos used a full blade rack containing 10 blade servers, each of which had four processors, making for a combined total of 40 (yes you read that right) processors, the team from Ben Franklin University used an array of Mac minis in their vehicle, 'Little Ben'. Stanford, though, kept things simple using just two computers with a single quad core processor in each. In fact, the motherboards were the same 975XBX2 motherboards we use in our test beds for graphics testing and the processors were just standard Intel Core 2 Duo Q6600s.
Similarly, while some teams had installed huge great after market air-conditioning systems to keep all the computing equipment cool, Stanford had few enough components that they could simply pipe the cold air from the car's air-conditioner into the racking in the boot of the car. Job done!
If you're wondering how all the cars were controlled, in the case of the Stanford team, it was again very simple. All the steering, braking and accelerating is controlled by computers in the production car anyway so all the team had to do was tap into the system and control them using the computers in the boot. This way the interior of the car looks almost completely normal and indeed the car could be driven quite normally by a human. In fact, all the teams in the final used similar methods and all the cars could be driven normally, but the Volskwagen Passat is the only car that has complete computer control over all its driving components.