Graphics for Serious Games RIST: Radiological Immersive Survey Training for two simultaneous users Steven Koepnick a , Roger V. Hoang b , Matthew R. Sgambati a , Daniel S. Coming a,Ã , Evan A. Suma c , William R. Sherman d a Center for Advanced Visualization, Computation and Modeling, Desert Research Institute, 2215 Raggio Parkway, Reno, NV 89512, United States b High Performance Computation and Visualization Laboratory, University of Nevada, Reno, 1664 N. Virginia Street Reno, NV 89557, United States c Institute for Creative Technologies, University of Southern California, 5318 McConnell Ave, Los Angeles, CA 90066, United States d Advanced Visualization Laboratory, Indiana University, 2711 E. 10th Street, Bloomington, IN 47408, United States article info Keywords: Virtual reality Computer-based training Human factors Collaborative virtual environment abstract National Guard Civil Support Teams (CST) respond to a variety of situations involving dangerous materials. Many of these situations can be safely simulated for training purposes in the real world. Radiological threats, however, are difficult to simulate due to the lack of materials that can mimic radiation sources without the danger of the real radiation. To address the need for a system to train CSTs to respond to radiological threats, we have developed the Radiological Immersive Survey Training (RIST) system. RIST simulates radiological threats from multiple sources using a realistic real-time shielding model based on ray casting and allows users to practice surveying the threat using simulated representations of the world and equipment. We have developed an after action review tool to allow a trainer to show trainees a recording of their survey and how they can improve. We also created a scenario design tool to allow the trainer to create complex environments with radiological threats. We developed novel multi-user interaction techniques to enable simultaneous training for two CST members in an immersive virtual environment. We also introduced a novel multi-perspective rendering technique for two users based on each user’s task rather than field of view. Finally, we conducted a preliminary user study with several pairs of expert users to measure user preferences and the effects of using this technique, in conjunction with how altering which user navigated, on user performance. CST survey teams from two states have now used the system for training. & 2010 Elsevier Ltd. All rights reserved. 1. Introduction Civil Support Teams (CST) survey chemical, biological, and radioactive threats in order to establish safe perimeters for the public and identify dangers. Training is critical to team member performance and safety while surveying a hazard. Training for CST tasks is a complex, coordinated team effort, and hazards, particularly radioactive, are currently difficult to safely simulate using traditional methods. Tate et al. have shown that training and mission rehearsal can benefit from virtual reality (VR) [1]. In particular, their research demonstrated that immersive systems can help users learn to perform visual and spatial tasks. Additionally, Chua et al. have shown that VR is effective for training to practice motor skills and perform movements [2]. We based large portions of our system’s functionality and interface on communications with several members of a CST [3]. Radiological surveys typically involve a team of two CST members who survey an area to identify radiation sources. The teams survey inward toward suspected sources from up to eight directions. Perimeters of increasing levels of danger are indicated by placing colored flags at specific radiation readings. Fig. 1 shows a survey team forming such perimeters. The team will use more complex surveying patterns in complex areas where approaching from many directions is not possible. Several existing training methods attempt to simulate this process. One method relies on survey meters that present preprogrammed radiation readings based on GPS location. Unfortunately, GPS readings are not sufficiently accurate for the task. Readings often fluctuate, providing inconsistent values. Another method requires a third person to consult a chart to determine the radiation reading the team should be observing at their current location. Inconsistency and inaccuracy are also problems with this method, as the third person can make mistakes while consulting the chart and must approximate the readings based on the team’s location. In a similar method, the team reads radiation readings from cards placed on the ground. Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/cag Computers & Graphics 0097-8493/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.cag.2010.09.008 Ã Corresponding author. E-mail addresses: steve.koepnick@dri.edu (S. Koepnick), rvhoang@gmail.com (R.V. Hoang), matthew.sgambati@dri.edu (M.R. Sgambati), suma@ict.usc.edu (E.A. Suma), shermanw@indiana.edu (W.R. Sherman), daniel.coming@dri.edu (D.S. Coming). Computers & Graphics 34 (2010) 665–676