Semi-automatic Reply Avatar for VR Training System with Adapted Scenario to Trainee’s Status

Abstract

For service industry, quality of customer service skills and customer satisfaction are very important topic. However, it is difficult to learn and master this kind of service skill without on-the-job-training (OJT). In this research, we propose a service VR simulator in which user can train man-to-man service by using VR technologies (Fig. 1). For service skill, it is important not only physical behavior and valval skills which can be learn by using manual, but also emotional control and sensing skills. Thus, to training the emotional skill on man-to-man service, we propose and construct preliminary service VR simulator to master emotional skills.

Our service VR simulator consist of mental/emotional sensing devices, estimating algorithm and intervention approaches. At first, to sensing mental/emotional state of trainer, we developed vital sensor attached HMD (Fig. 2). By sensing vital signal, such as heart beat and respiration, we can estimate mental/emotional state of trainer during VR training. Next, we also develop mental/emotional intervention which can induce emotion of users. For example, we can alleviate user’s tension with altered auditory feedback. Finally, our proposed system can control difficulty of training according to trainer’s mental state and skills. Combining these mental/emotional sensing and feedback technologies enables effective training of emotional control and expression skills.

In addition, by utilizing advantage of the characteristics of VR, our system can be expected to be more effective than training conducted on actual sites. For example, by switching to customer’s viewpoint from trainer’s own viewpoint, trainer can find and understand his/her problem easily. Also, by replaying and rewinding man-to-man service simulation, trainer can try-and-error and find solution during training.