Research driving simulators are used by Universities, the car industry and research institutes for doing behavioural research into car driving. One of the most well known simulators is the research simulator in Iowa. Another important one is the VTI research simulator. While these have cost millions of dollars to develop and have expensive motion platforms, most commercially available research simulators are a bit more limited in their hardware. The most important part of a research simulator functionality is the software that enables researchers to do the research they want.
A researcher needs tools to convert their research questions into scenarios that run in the simulator. These can have the following functionality:
- define the variables that must be measured, and also whether these should be sampled with a fixed sample frequency or should alread be processed in real time during the experimental trials. Examples of the first type are pedal positions, longitudinal and lateral velocity, acceleration, lateral position, distance or time to intersection, etc. Examples of the second type are standard deviation of lateral position, minimum of time to collision, time to line crossing, brake reactio time, steer reaction time, etc.
- define the traffic situations in a way that is the same for all subjects.
- define the virtual environment he experiment is situated in, for example a highway, a motorway, a village, etc
In some research simulators, the traffic scenarios are static. This means that something occurs when the driver is at a certain position. Other research simulators allow for dynamic scenarios: this means that something occurrs when the driver has reached a time-related criterion. For example, when the time to intersection is less than 5 seconds, the traffic light turns yellow. Dynamic scenarios are better because they ensure that task load is the same for all subjects.
Some simulators have a script language interface to define the scenarios, while others use a graphical editor or a configuration file to define the scenarios. In general, script languages are a bit more difficult to learn but the are much more flexible and versatile that other interfaces.
Another important difference is the range of variables that can be measured. Advanced driving simulators have a wider range that include time based measures such as time to collision and time to line crossing. They also have the functionality to let users define their own type of variables.
These are a few differences between simulators the user must check before they choose a research driving simulator.
If you are interested in good examples of behavioural studies into car driving that have been performed in a driving simulator you should definitely check out this website on driving simulators in traffic psychology. It also presents an overview of models of driving behaviour.