Scholarlink

<p>Industry 4.0 technologies are revolutionizing maintenance practices by enabling the use of smart, data-driven strategies that boost equipment reliability and minimize downtime. This contribution illustrates the practical application of a recently developed Industry 4.0 deployment model for maintenance optimization within a palm oil refining company. A key part of this application is the development of a digital twin prototype for a critical polishing filter, designed to support real-time process monitoring and predictive maintenance strategies at the same time, the remaining useful life (RUL) of the filter’s internal sleeves is predicted using advanced machine learning methods. In particular, the Extreme Gradient Boosting (XGBoost) algorithm is applied to historical process data to predict sleeve degradation cycles. Whereas the digital twin provides a dynamic representation of the filtration unit, predictive models serve as an independent analytical lens for maintenance planning. This blended approach demonstrates the power of Industry 4.0 technologies to improve maintenance decision-making in the agro-industrial sector and provides valuable guidance for organizations looking to advance their digital transformation.</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

Application of an industry 4.0 implementation model: digital twin-based predictive maintenance of a polish filter in palm oil refining

  • 期刊论文

  • 发表日期：

    2026年1月10日

查看全文
  • Safaa Essalih,
  • Zineb El Haouat,
  • Mohamed Ramadany,
  • Fatima Bennouna,
  • Driss Amegouz

摘要

Industry 4.0 technologies are revolutionizing maintenance practices by enabling the use of smart, data-driven strategies that boost equipment reliability and minimize downtime. This contribution illustrates the practical application of a recently developed Industry 4.0 deployment model for maintenance optimization within a palm oil refining company. A key part of this application is the development of a digital twin prototype for a critical polishing filter, designed to support real-time process monitoring and predictive maintenance strategies at the same time, the remaining useful life (RUL) of the filter’s internal sleeves is predicted using advanced machine learning methods. In particular, the Extreme Gradient Boosting (XGBoost) algorithm is applied to historical process data to predict sleeve degradation cycles. Whereas the digital twin provides a dynamic representation of the filtration unit, predictive models serve as an independent analytical lens for maintenance planning. This blended approach demonstrates the power of Industry 4.0 technologies to improve maintenance decision-making in the agro-industrial sector and provides valuable guidance for organizations looking to advance their digital transformation.