About this Research Topic
Currently, most applications of LfD in industrial manufacturing use kinesthetic teaching, that is, the human operator teaches the robot by physically moving the actuator, demonstrating the desired movements and actions. However, for learning robots to be able to live to their potential their capabilities should go beyond the simple and mindless replication of actions. Ideally, they would actually understand the task at hand, in terms of its end-goals, the causal relations between its constituting sub-tasks and the relevant states of the surrounding environment. Not only this could greatly foster the applicability of LfD in manufacturing scenarios, but it could also enable true collaboration with a human operator, since advanced task understanding is a critical steppingstone to enable joint action.
In the scope of the Cyber-Physical Human System paradigm, this Research Topic aims to provide a forum for engineers, data scientists, researchers, and practitioners to present innovative research related to algorithms and methods supporting robot LfD systems and frameworks concerning intelligent manufacturing processes. It covers all aspects of the field, from advances in more traditional low-level trajectory demonstration to high-level task understanding and goal inference that "mimics" human intelligent learning, allowing to improve synergistic human-robot collaboration. Review articles and innovative works on performance evaluation and benchmark datasets are also solicited. Potential topics of interest include, but are not limited to the following:
● Learning by observation
● One-shot learning
● Collaborative robot learning
● Active learning for manufacturing
● Kinesthetic teaching
● Efficient data labeling
● Semantic scene understanding
● Transfer learning algorithms in smart manufacturing
● Knowledge generalization
● Robotics-based digital twins
● Human-robot interfaces
Keywords: Robot learning, active learning, computer vision, machine learning, human-robot interfaces
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.