Grupo DIANA

The primary aim of this study is to evaluate the effectiveness of a Virtual Reality (VR)-simulated system in training fine motor skills that can be transferred to the performance of manual tasks in the real world. The VR system presented in this thesis enables training following fundamental methods such as part-task and whole-task training, in realistic simulations which are enhanced through haptic interaction. A fundamental advantage of using virtual reality in training is the ability to provide specific types of augmented feedback which cannot be provided in the real world. The work presented in this thesis fits within the scope of ManuVAR (Manual work support throughout system lifecycle by exploiting Virtual and Augmented Reality), a European Union (EU) funded Seventh framework programme project which, among other things, aimed to support motor skill training in high value high knowledge manual work by using Virtual Reality technologies. A case study in industrial maintenance is presented: the metallographic replica, a nondestructive inspection technique that requires fine grinding and polishing of the inspected area. The motor skills required for the performance of these tasks must be particularly accurate. However, those motor skills consist of a tacit knowledge which is hard to transfer from experts to trainees. This thesis focuses on the design and the evaluation of a VR training system which aims to supplement the motor skill training traditionally carried out for the performance of fine grinding and polishing tasks. The VR training system was designed on the basis of functional and customer requirement analyses which enabled defining the functionalities that allow solving the issues that arise when training in the real world. Two experimental studies were designed to investigate whether a training program inspired by part-task and whole-task training methods, along with the provision of augmented feedback, enabled training the motor skills that are relevant for the performance of fine grinding and polishing tasks. The first experimental study explored the effectiveness of part-task training on the performance of a polishing task in a virtual environment. The second study evaluated the effectiveness of the complete training program for both tasks and investigated the capability of the VR training system to discriminate between several levels of expertise. The outcomes of the experimental studies show the effectiveness of the training carried out on the VR training system, showing meaningful accuracy improvements throughout the performance of motor skills. This proves the internal validity of the proposed training. Moreover, the construct validity of the system is also suggested through the discrimination between expert and non-expert operators. On the basis of these findings, the external validity of the VR training system to train the fine motor skills that are relevant for the performance of fine grinding and polishing tasks in real operating environments can be established. This work supports the hypothesis that VR enhanced with haptic force feedback can be useful for training fine motor skills, complementing the traditional training, which is carried out in real operating environments.