Bringing a Natural Language-enabled Virtual Assistant to Industrial Mobile Robots for Learning, Training and Assistance of Manufacturing Tasks

Abstract

Nowadays, industrial companies want to enhance their Industry 4.0 competencies. Therefore, they need to help employees master state-of-the-art technologies and gain the necessary knowledge to stay relevant and competitive. As a result, there is a global demand for learning and training tools that assist the employees at all levels. In this paper, we propose a natural language-enabled virtual assistant (VA) integrated with an industrial mobile manipulator to fulfill this target in manufacturing tasks. The latest Learning, Training, Assistance - Formats, Issues, Tools (LTA-FIT) model is leveraged to guide the design and development of a pilot version of the VA. To validate its performance, three manufacturing scenarios are analyzed based on the learning, training, and assistance phases, respectively. In our system, the human-robot interaction is achieved through conversation and a dashboard implemented as a web application. This intuitive interaction enables operators of all levels to control a industrial mobile manipulator easier and use it as a complementary tool for developing their competencies. The pilot experiments show that the proposed VA is able to respond to operator commands flexibly within the LTA-FIT model.