All About Partial Virtual Environment

The way STISIM Drive creates simulation scenarios is different from other driving simulators. Unlike most simulation programs, the STISIM Drive method is best described as partial virtual environment generation. That means that as a driver is going through a simulation, only a portion of the virtual world is displayed. This is the basis for how the virtual roadway environment is generated, and for how the driving scenarios are designed using the Scenario Definition Language (SDL).

Figure A (left) Coordinate map-based VE; Figure B (right) Partial VE generation.


Let’s break it down a bit more with the following example. The figures above both show a vehicle approaching an intersection. In other simulation programs that use a coordinate map-based system (Fig A), the direction the driver takes will determine what he sees in this virtual world. Whichever way he goes (continues to go straight or turns right/left) will send him into a different section of the scenario.

If this was an STISIM Drive simulation (Fig B), no matter which way the driver goes (left/right/straight), he will go into the same section. This is because of the way scenario events are defined in the SDL.

Events in a scenario are “turned on” or “made live” based only on the driver’s longitudinal distance within the virtual environment. Since an event (in this example, Section B) can be specified to occur after a driver reaches a specific distance on the virtual road, until that distance is traveled, the event does not exist.

As the driver travels through the virtual environment, he continues to accumulate longitudinal distance. Once the driver reaches the specified distance in the scenario, then the event will be triggered to appear according to its parameters.

Moving into different sections when turning is intuitive and not usually a problem when designing virtual environments in relation to a coordinate map context. But if the goal of the scenario is to measure driver behavior in response to a particular event (e.g. a pedestrian crossing the street, or another vehicle pulling out in Section B), then scenario design becomes problematic.

In other simulation programs, unintended turning (meaning the researchers didn’t intend for a driver to turn a certain way) could cause unintended consequences when the boundaries of the virtual environment are exceeded, or you get ordering effects in your data because the driver did not follow the intended path.

As far as we’re concerned, there are many advantages to the way STISIM Drive generates scenarios – especially for experimental driving research. The main focus of driving simulation is to measure a driver’s behavior during specific situations, and therefore you need to insure that the driver encounters the desired situations. In STISIM Drive, the roadways are presented serially, which means all drivers who drive a scenario can potentially experience the same things, regardless of turning behavior.

Drivers won’t get disoriented or lost in the virtual environment. They can’t make a wrong turn and they won’t “break” the scenario as they drive it. Plus, drivers will experience the roadway events as the researcher intend them to. And the counterbalancing of scenario events or whole roadway sections can then be easily accomplished to control for order effects. Best of all, this method can reduce the design requirements and development times for scenarios.

However, there are some limitations to STISIM Drive’s partial VE generation. One of the main limitations is the inability to simulate specific geography relative to a coordinate map-based system. This means that studies involving simulation with GPS mapping and navigational tasks are problematic. Also, there is the possibility of non-realistic route corrections if a driver makes a navigational error. For instance, if a driver makes a wrong turn and instinctively corrects himself by making a U-turn, an issue could occur since the program provides only a limited distance of backtracking.

Some of our customers have gotten around this by modifying general program settings and adding a single, elaborate, large scale 3D city model. But it should be noted that instituting these applications would require considerable 3D modeling resources since our system was not conceptually designed in this manner.

Bottom line, driving simulation systems handle virtual environments, and how drivers act in those environments, in slightly different ways. STISIM Drive uses partial virtual environments to give researchers better control over the events in a VE so they can better measure driver behaviors.