Heuristic guidelines and experimental evaluation of effective augmented-reality based instructions for maintenance in nuclear power plants

Abstract

As industrial plants and factories age, their maintenance requirements increase. Because maintenance mistakes directly increase the operating costs of a power plant, maintenance quality is significant concern to plant management. By law, all personnel working with nuclear technology must be re-trained every three years in Korea; however, as the statistical data show, the number of shutdown accidents at nuclear power plants (NPPs) due to maintenance failure is still high and needs to be reduced. Industries have started to adopt various technologies to increase the speed and accuracy of maintenance. Among those technologies, augmented reality (AR) is the latest multimedia presentation technology to be applied to plant maintenance, and it offers superior intuitiveness and user interactivity over other conventional multimedia. This empirical study aims to measure the optimum amounts of information to be delivered at a time and to identify what types of information enhance the learning ability of novices and to suggest heuristic guidelines by which to make effective AR training instructions. In the first experiment, the optimum amount of information in an AR learning environment for novices was found to be 4-5 pieces of information in a chunk by comparing results between a pre-test and an after-test. This result implies that intentionally made chunks help novices learn more effectively. In the second experiment, the AR training instruction based on the suggested heuristic guidelines was slightly more effective than other AR training instructions. Maintenance in nuclear power plants can be more reliable and accurate by training through AR training instruction based on the suggested heuristic guidelines. (c) 2010 Elsevier B.V. All rights reserved.