Towards Remote Expert Supported Autonomous Assistant Robots in Shopping Environments

Abstract

Autonomous robots that assist customers during their shopping trips may become ubiquitous in the near future. However, these systems are limited in their ability to appropriately reply to all of a customer's requests. Therefore, a dedicated human expert has to remotely "take over" and resolve the customer's issue to warrant a satisfying experience. As a result, a Transfer of Control from a machine to a human takes place. This creates the problem that the expert needs to quickly review what has happened prior to the take-over to provide optimal assistance. To address this, we designed and implemented an interactive summary concept allowing experts to quickly filter and review the most relevant information to solve the given issue. Finally, we provide an example use case illustrating how a remote expert may solve a request using our system.