Evaluating an Augmented Reality-Based Partially Assisted Approach to Remote Assistance in Heterogeneous Robotic Applications

Abstract

Among the countless applications of Augmented Reality (AR) in the industry, remote assistance represents one of the most prominent and widely studied use cases. Recently, the way in which assistance can be delivered started to evolve, unleashing the full potential of such technology. New methodologies have been proposed able to foster operators' autonomy and reduce under-utilization of skilled human resources. This paper studies the effectiveness of a recently proposed approach to AR-based remote assistance, referred to as partially assisted, which differs from the traditional step-by-step guidance in the way the AR hints are conveyed by the expert to the operator. The suitability of this approach has been proved already for a number of simple industrial tasks, but a comprehensive study has yet to be performed for validating its effectiveness in complex use cases. This paper addresses this lack by considering as a case study the mastering of a robotic manipulator, a procedure involving a number of heterogeneous operations. The performance of the partially assisted approach is compared with step-by-step guidance based on both objective and subjective metrics. Results showed that the former approach could be particularly effective in reducing the time investment for the expert, allowing the operator to autonomously complete the assigned task in a time comparable to traditional assistance with a negligible need for further support.