A Practical Roadmap to Learning from Demonstration for Robotic Manipulators in Manufacturing

TL;DR

The paper tackles the challenge of applying Learning from Demonstration in manufacturing by presenting a practical, questionnaire-style roadmap for robotic manipulators. It articulates four guiding questions—What to Demonstrate, How to Demonstrate, How to Learn, and How to Refine—and dissects them through task scope, demonstration mechanisms, learning spaces, and learning methods. The authors synthesize methodological options (e.g., MP, DMP, RL/IRL, GP, GMM, ProMP) and offer concrete guidance on when and how to use each, including context-dependent demonstrations and safety considerations. They further outline refinement strategies to improve learning performance, accuracy, robustness, and human-robot interaction safety, aiming to bridge research and industry practice. Overall, the work provides a structured, actionable framework enabling practitioners to adopt LfD for customizable manufacturing tasks with a focus on manipulators and practical deployment.

Abstract

This paper provides a structured and practical roadmap for practitioners to integrate Learning from Demonstration (LfD ) into manufacturing tasks, with a specific focus on industrial manipulators. Motivated by the paradigm shift from mass production to mass customization, it is crucial to have an easy-to-follow roadmap for practitioners with moderate expertise, to transform existing robotic processes to customizable LfD-based solutions. To realize this transformation, we devise the key questions of "What to Demonstrate", "How to Demonstrate", "How to Learn", and "How to Refine". To follow through these questions, our comprehensive guide offers a questionnaire-style approach, highlighting key steps from problem definition to solution refinement. The paper equips both researchers and industry professionals with actionable insights to deploy LfD-based solutions effectively. By tailoring the refinement criteria to manufacturing settings, the paper addresses related challenges and strategies for enhancing LfD performance in manufacturing contexts.

Figures (6)

• Figure 1: Overview of our proposed roadmap for LfD implementation.
• Figure 2: Illustration of how subtasks and task hierarchy comprise a full task.
• Figure 3: Comparison of motion-based versus contact-based task. On the left, the pick-and-place task has a structured and predictable interaction with the environment, while the insertion task on the right needs to deal with the contact resulting from tight tolerances to successfully perform the task.
• Figure 4: Illustration of compliant versus non-compliant behavior against the environment. The black line represents the environment surface, the red path represents a non-compliant behavior, and the blue path represents the compliant behavior against the environment surface. In Impedance Control, the end-effector is modeled as a spring-damper system.
• Figure 5: Illustrative examples of the main demonstration approaches.