TESTING HUMAN ERRORS IN VIRTUAL REALITY TRAINING

  • 1Brno University of Technology, Dept. of Production Systems and Virtual Reality, Brno, CZ
  • 2Brno University of Technology, Research Center of Automatic Manipulation, Brno, CZ
  • 3Brno University of Technology, Production Machines and Robotics Division, Brno, CZ
  • 4Brno University of Technology, Dept. of Quality, Reliability and Safety, Brno, CZ
  • 5Kovosvit MAS, Sezimovo Usti, CZ

Abstract

The emergency training of industrial process plant operators is one of the most widely used tools to increase the reliability of human factors to handle an emergency situation. However, the preparation and operation of full-fledged simulators and trainers is very expensive and, therefore, virtual environment tools are used. A question that has not yet been answered is: Can virtual reality match the reliability of other methods of operation and is the same training in virtual reality effective? The experiment was carried out in the three-walled virtual CAVE, with virtual reality glasses, with a computer, a tablet, and a real control panel. Visual stimuli were displayed on the screen of the virtual monitor (green, yellow, and red); auditory stimuli were pure tones with frequencies of 250, 1000 and 4000 Hz. The conclusion should explicitly state if the hypothesis defined for the research has been confirmed and there are significant differences in terms of interface type. Training in virtual reality induces lower operator reliability, but in specific conditions (visual stimuli, virtual reality glasses) can match the reliability of other methods of operation and can be effective.

Recommended articles

CALIBRATION OF THE ROBOTIC ARM WITH CORRECTIONS USING LOCAL LINEAR NEURO-FUZZY MODELS

MICHAEL VALASEK, ZDENEK NEUSSER, ZBYNEK SIKA, JAN PELIKAN, PETR BENES, JAN HLADIK, MARTIN NECAS, JIRI SVEDA
Keywords: Robot calibration | Local linear model | Neuro-fuzzy model | parameter identification | Accuracy of machine positioning