Scholarlink

<p>The advent of sixth-generation (6G) wireless communication systems will transform the control of industrial robots with Ultra-Reliable Low-Latency Communication (URLLC) for mission-critical automation. This study aims to enhance the robotics systems of smart manufacturing by predicting performance metrics of two key network indicators, packet error rate and end-to-end latency, which affect the stability and responsiveness of real-time industrial manipulators and collaborative robots (cobots). To accomplish this, a machine learning-based predictive system is established using the following techniques: Linear Regression, Random Forests, Neural Networks, and Extreme Gradient Boosting (XGBoost). The models are trained and tested in dynamic, realistic industrial communication environments to account for resource scheduling and wireless communication fluctuations. The results show that the XGBoost model achieves the highest prediction accuracy, outperforming baseline approaches, making it highly suitable for real-time industrial applications. In addition, Shapley Additive Explanations (SHAP) are used to enhance interpretability and pinpoint the parameters with the greatest impact on system performance. The explainability analysis shows that the dominant factors in fluctuations in latency and reliability at the transmission level are the latency allocation ratio, resource allocation, and the inverse of signal strength. Overall, this work validates the effectiveness of combining high-performance machine learning with explainable AI to optimize 6G-enabled industrial robotic communication systems. The findings provide actionable recommendations for smart resource allocation and lay a foundation for edge-assisted automation in future smart factories.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Explainable machine learning framework for predicting latency and error rate in 6G-enabled industrial robotic systems

  • 期刊论文

  • 发表日期：

    2026年5月13日

查看全文
  • Nagappan Partheeban,
  • Sivakumar Premkumar,
  • Tamilmani Sundari Arthi,
  • Yu-Chen Hu

摘要

The advent of sixth-generation (6G) wireless communication systems will transform the control of industrial robots with Ultra-Reliable Low-Latency Communication (URLLC) for mission-critical automation. This study aims to enhance the robotics systems of smart manufacturing by predicting performance metrics of two key network indicators, packet error rate and end-to-end latency, which affect the stability and responsiveness of real-time industrial manipulators and collaborative robots (cobots). To accomplish this, a machine learning-based predictive system is established using the following techniques: Linear Regression, Random Forests, Neural Networks, and Extreme Gradient Boosting (XGBoost). The models are trained and tested in dynamic, realistic industrial communication environments to account for resource scheduling and wireless communication fluctuations. The results show that the XGBoost model achieves the highest prediction accuracy, outperforming baseline approaches, making it highly suitable for real-time industrial applications. In addition, Shapley Additive Explanations (SHAP) are used to enhance interpretability and pinpoint the parameters with the greatest impact on system performance. The explainability analysis shows that the dominant factors in fluctuations in latency and reliability at the transmission level are the latency allocation ratio, resource allocation, and the inverse of signal strength. Overall, this work validates the effectiveness of combining high-performance machine learning with explainable AI to optimize 6G-enabled industrial robotic communication systems. The findings provide actionable recommendations for smart resource allocation and lay a foundation for edge-assisted automation in future smart factories.