Making the usage of guidance systems in pedestrian infrastructures measurable using the virtual environment DAVE
This paper presents the development and partial validation of a cave automatic virtual environment (CAVE), which is designed to make the reactions of pedestrians to guidance information (even such that are not technically feasible with current technology) in pedestrian infrastructures (such as airports, train stations or subway stations) measurable. The navigation is designed to be intuitive and easy to learn. It uses the Microsoft Kinect to obtain information on the user's movement. The user walks in place to move forward in the virtual world and turns her shoulders to invoke rotations in the virtual world to make turns. The virtual world includes simulated pedestrians to enhance the immersion and is equipped with a number of sensors that allow for a multi-method measuring of users. After the implementation of hands free steering two case studies are used to provide first evidence with respect to the possible answers that the research infrastructure is capable of delivering. The validity of the model for steering has been explored using a case study involving parallel test groups that expose individuals to wayfinding exercises in both the real world and the corresponding virtual world. Our results show that the objective distances and times in the real and the virtual worlds, as well as perceptions of distances, times and directions, do not differ statistically significantly. This provides a partial validation of the model for steering. In a second larger case study the hypothesis was tested that using the virtual environment test persons are able to find their way also in complex multi-level infrastructures with only limited learning requirements. Additionally, we tested the hypothesis that also in this setting realistic paths (including elevator and escalator usage) are taken by the test persons while observing realistic average velocities over longer trips. We find that our hypotheses have not been rejected by the data. Therefore, this environment is a useful tool for the design of guiding systems for large pedestrian infrastructures.