Adaptive Human-Robot Collaboration in Industrial Assembly: Augmented Reality-Supported Dynamic Task Allocation with Intuitive Process Planning

Abstract

As industrial assembly faces increasing demands for cost-efficiency and flexibility, human-robot collaboration (HRC) emerges as a promising solution for small to medium production volumes. To address practical challenges in HRC implementation – such as effective task distribution, trust, and workers’ information needs – this paper introduces a system that uses augmented reality (AR) to enhance adaptive HRC. The system intuitively provides essential information to workers and dynamically adjusts task allocation based on individual performance. It consists of five core components: an AR-integrated situation recognition system, situation-aware robot path planning, dynamic task allocation, AR visualization of work information, and a no-code software interface with a digital twin for process creation and monitoring. The paper offers a comprehensive overview of the system architecture, the user interfaces for process creation and AR visualization, and details a novel approach for optimization-based dynamic task allocation. Improvements in efficiency and collaboration are demonstrated through a practical assembly scenario at a laboratory workstation for collaborative assembly.