Intention recognition-based human-machine interaction for mixed flow assembly

doi:10.1016/j.jmsy.2023.11.021

Intention recognition-based human-machine interaction for mixed flow assembly

2024-02-01

7

PRE

AI

DOI：10.1016/j.jmsy.2023.11.021

Original

Original request for help

Save

Abstract

En 中文

Mixed-flow assembly provides a highly adaptable solution for production processes, catering to the extensive customization requirements in product manufacturing. Human-machine interaction (HMI) can significantly enhance the efficiency, flexibility, and responsiveness of mixed-flow production lines by enabling operators to collaborate with machines. However, current HMI research primarily focuses on temporal or spatial interaction, where tasks are completed independently by the operator and machine in a predetermined sequence, with emphasis on collision avoidance in a shared workspace. Research on spatial-temporal HMI is currently lacking. This paper addresses this gap through three comprehensive steps. Firstly, we develop an intention recognition model based on Gaussian mixture distribution to effectively identify the user's operational intention. Subsequently, we design a hybrid industrial communication architecture to facilitate intent sharing among various production elements. Finally, an empirical study is conducted to assess the impact of different forms of HMI on assembly efficiency and error rates. Our experimental results demonstrate that spatial- temporal HMI can enhance assembly efficiency by approximately 15% and reduce the defective rate. In future studies, the proposed spatial-temporal cooperation platform could integrate additional perception technologies to alleviate assembly fatigue and improve operators' well-being.

Keywords:

Human-machine interaction

Mixed-flow assembly

Industrial communication

Intention recognition

Intention recognition-based human-machine interaction for mixed flow assembly

Abstract

Journal

Organization