Multi-modal interfaces for natural Human-Robot Interaction

摘要

Recent years, manufacturing aims on increasing flexibility while maintaining productivity for satisfying emerging market needs for higher product customization. Human Robot Collaboration (HRC) is able to bring about this balance by combining the benefits of manual assembly and robotic automation. When introducing a hybrid concept, safety and human acceptance are of vital importance for achieving implementation. Fenceless coexistence may lead to discomfort of operators especially in cases where close Human Robot Interaction (HRI) occurs. This work aims at designing and implementing a natural Human-System and System-Human interaction framework that enables seamless interaction between operators and their “robot colleagues”. This natural interaction will strengthen hybrid implementation through increased: a) operator’s and system’s awareness, b) operator’s trust to the system, and through the decrease of: a) human errors and b) safety incidents. The overall architecture of the proposed system makes it scalable, flexible, and applicable in different collaborative scenarios by enabling the connectivity of multiple interfaces with customizable environments according to operator’s needs. The performance of the system is evaluated on a scenario originating from the automotive industry proving that an intuitive interaction framework can increase acceptance and performance of both robots and operators.