Validation of a system for the simulation of interactive virtual environments

Status:

completed 05/2011

Aims:

The conditions and processes of real working environments can be simulated by means of virtual reality (VR) in a VR laboratory. This enables interaction between human beings and technical installations to be analysed, designed and evaluated under realistic conditions. In addition, new developments in technical installations, machine controls or safety facilities or major changes to them can be modelled and studied in a VR laboratory relatively quickly and cost-effectively. Conditions and processes that do not yet exist or that could present a hazard can also be simulated and studied in this way.
The project's primary purpose is to evaluate the realism of VR. The SUTAVE VR laboratory at the Institute for Occupational Safety and Health of the German Social Accident Insurance (IFA) was to be validated in order to yield sample information on the correspondence between the results obtained by the laboratory with those of real working processes, and scope of transfer of the former to the latter.

Activities/Methods:

A literature survey was first conducted in order to gather findings concerning the realism of virtual environments and methods for comparing real and virtual environments.
Development work was also to be performed in order to align the real and virtual environments with each other. In the process, functionality (such as display stereoscopy) of the VR laboratory was examined and optimized. Methods for the recording and analysis of data of human-machine interaction were also developed, and integrated into the SUTAVE laboratory.
Psychological concepts for human information processing served as the theoretical basis for this project. These concepts place the emphasis upon the perception, processing and implementation of information as major components within the work process which are equally crucial for evaluation of the interaction between human beings and (virtual) work environments. One part of the evaluation studies addressed the acquisition of information. For this purpose, the quality of presentation of impact variables in work processes (such as luminance, depth/colour perception) was studied in the real and in the virtual environment. The other part of the evaluation studies also considered processing and implementation of information. Therefore, effects of human-system interaction and in particular of the human-robot interaction upon behaviour dimensions were studied.

Results:

The results of the evaluation studies show that the VR laboratory at the IFA permits realistic simulation of a range of work conditions and processes. Demonstrations and empirical studies can be conducted for OSH purposes. The IFA's VR laboratory is designed for the use of VR for OSH purposes as documented in the national and international literature, and presents potential for extending this use.
One of the empirical studies concerned the quality of presentation of impact variables in work processes. Measurements relating to the luminance revealed a lower level in the virtual robot cell compared to the real environment; contrast levels in the two work areas were however on the whole comparable. The systematic, experimental comparisons of depth perception confirmed the discrete observations made in the past by visitor groups relating to the realistic presentation of work processes in VR. However, studies of the depth perception in the virtual and real robot cell also showed that estimates of size ratios between objects in virtual areas exhibit somewhat higher error rates. The colour perception in VR was impaired to a greater degree when compared to the real environment only when a large number of secondary colours were used. This phenomenon, which is also known in the literature, is to be dealt with in future VR studies by adjustment of the presentation of information.
A study into the effects of the human-system interaction upon the behaviour also made use of the robot cell simulated in the IFA's VR laboratory on a scale of 1:1. Completion of the task in the virtual robot cell provided the participants with a high degree of immersion and presence, comparable with the results obtained in international studies. During the studies, symptoms of simulation sickness remained at negligible levels, comparable with international reference values. Different intensities of human-robot interaction were reflected by all three dimensions of human behaviour (psychophysiological responses, sensation, human task performance).
By means of these studies, it was possible to document the suitability of the IFA's SUTAVE laboratory for the analysis, design and evaluation of human-system interaction. As a simulation and method, VR will not be able to replace field studies, but may reduce their necessary scale. The results of this project show that VR can be used to perform valid studies for OSH purposes.

Last Update: