An intuitive robot manipulation interface based on mixed reality and multi-point mapping for assembly applications

In Industry 5.0, high-precision human–robot collaborative assembly requires intuitive interfaces that minimize cognitive load. However, traditional 6-DoF control interfaces are hindered by kinematic coupling, as the control coordinate system (CCS) is rigidly fixed to the robot’s end-effector (EEF), forcing operators to perform complex compensatory movements. To address this, we propose the Multi-Point Mapping Interface (MPMI), a Mixed Reality (MR) strategy that dynamically decouples the CCS from the EEF, allowing operators to align control with task-relevant geometric features. A single-factor within-subjects user study ((Formula presented) ) was conducted to validate the system against a traditional Single-Point Mapping Interface (SPMI). Experimental results demonstrate that MPMI significantly reduced task completion time by 14.6% ((Formula presented) ) and improved input efficiency, as indicated by a 15.0% reduction in cumulative operator input pose change ((Formula presented) ). Furthermore, subjective assessments confirmed a significant decrease in NASA-TLX cognitive load ((Formula presented) ) and superior usability of the system. These findings empirically validate the principle of dynamic control origin decoupling as a critical methodology for enhancing efficiency in complex robotic assembly, providing a foundation for future cognitive-optimized HRC systems.