Friday 14 May 2010

Driving without moving

Earlier this week, Phil and I went on a scouting mission to check out some driving simulators. We first went to visit Cruden in Amsterdam, who manufacture high-end driving simulators with very accurate motion cues. The platforms they use are very impressive pieces of engineering (see photo) and the degree of realism is astonishing. I got to drive a simulated Ferrari around a racing track, and while I haven't driven one in real life, the feeling of accelerating or hitting the break was very convincing - the clutch felt like a bull was hitting the car every time I changed gear.

Frank at Cruden was kind enough to also organise a quick meeting with Riener Haapee from Delft University. We learnt a lot about the research they are doing, which is closely related to what we want to do with the driving simulator. They have been using different types of car simulators to work on biomechanics and human-machine interaction in the car. (Another interesting bit of information was that apparently, every driving school in the Netherlands uses a (simple) simulator.) One of the projects at Delft University is DriveOps, which looks really interesting. Riener also pointed us some collaborators they are working with, including Noldus, a company creating behavioural models, and to SmartEye, a company creating eye trackers specifically for cars.

The next day, we went on to visit Kai Foerst (from Dr. Foerst GmbH) near Cologne, who make a different type of simulator, which is aimed much more at training drivers. While the motion cues are much less realistic than in the Cruden simulator, the actual software provides a great environment to quickly create and test driving scenarios in quite realistic everyday environments (see photo below).

One interesting aspect we discussed at length is the trade-off between realism of the motion cues and the risk of motion sickness, which is closely linked to the overall question of how realistic does the simulation need to be in order to be effective? Obviously, if the motion cues, the visuals and the interaction are inconsistent, then the degree of inconsistency will increase the likelihood of motion sickness. Another good predictor of motion sickness seems to be age: the older people are, the more likely they are to experience it. Younger people (e.g. those training to get a driver license) rarely suffer from it - we thought that maybe exposure to 3D video games is a factor in this case. However, an interesting question is, if this is indeed true, would that compromise the effectiveness of the simulator as people would (subconsciously) treat it more as a video game rather than as a simulation of the real world.

No comments: