Learning from Demonstrations in Human–Robot Collaborative Scenarios: A Survey

Human–Robot Collaboration (HRC) is an interdisciplinary research area that has gained attention within the smart manufacturing context. To address changes processes, HRC seeks to combine impressive physical capabilities of robots with cognitive abilities humans design tasks high efficiency, repeatability, and adaptability. During implementation cell, a key activity robot programming takes into account not only restrictions working space, but also human interactions. One most promising techniques so-called Learning from Demonstration (LfD), this approach based on collection learning algorithms, inspired by how imitate behaviors learn acquire new skills. In way, task could be simplified provided shop floor operator. The aim work present survey technique, emphasis collaborative scenarios rather than just isolated task. literature was classified analyzed on: main algorithms employed for Skill/Task learning, level participation during whole LfD process. Our analysis shows intervention been poorly explored, its implications have carefully considered. Among different methods data acquisition, prevalent method guidance. Regarding modeling, such as Dynamic Movement Primitives Semantic were preferred low-level high-level solving, respectively. This paper aims provide guidance insights researchers looking introduction in robotics context identify opportunities.