Scholarlink

<p>Artificial intelligence and machine learning are at the heart of the Industry 4.0 revolution, driving the digital transformation and modernization of manufacturing processes. Through the Industrial Internet of Things (IIoT), interconnected production nodes enable continuous monitoring and real-time processing of large data streams, allowing autonomous decision-making without human intervention. These technologies help eliminate inefficiencies, reduce production rejects, and accelerate design cycles. To maximize productivity, predictive maintenance combines IIoT data and machine learning to forecast the precise times when equipment and tools require servicing, minimizing downtime and optimizing scheduling. This paper introduced a potentially promising framework for predictive maintenance in the field of machine milling. The framework is based on a modified firefly algorithm, and it was tested and verified on two publicly available datasets. The aim of the framework is to monitor the wear of the milling tools and signal when replacement is required, thereby avoiding tool breakdown, saving time, and thus improving the productivity of the machine. Due to the intensive nature of optimization and data processing tasks, which involve large-scale multidimensional datasets and iterative model refinement, the proposed method benefits from high-performance computing resources. The algorithm’s design supports parallel and distributed execution, enabling real-time processing of sensor data and rapid decision-making required for autonomous manufacturing systems.</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

Exploring the applicability of decision trees and deep neural networks optimized by metaheuristics for predictive maintenance in milling

  • 期刊论文

  • 发表日期：

    2025年11月27日

查看全文
  • Aleksandra Bozovic,
  • Luka Jovanovic,
  • Milos Dobrojevic,
  • Milos Antonijevic,
  • Nebojsa Bacanin,
  • Eleonora Desnica,
  • Vladimir Simic,
  • Miodrag Zivkovic

摘要

Artificial intelligence and machine learning are at the heart of the Industry 4.0 revolution, driving the digital transformation and modernization of manufacturing processes. Through the Industrial Internet of Things (IIoT), interconnected production nodes enable continuous monitoring and real-time processing of large data streams, allowing autonomous decision-making without human intervention. These technologies help eliminate inefficiencies, reduce production rejects, and accelerate design cycles. To maximize productivity, predictive maintenance combines IIoT data and machine learning to forecast the precise times when equipment and tools require servicing, minimizing downtime and optimizing scheduling. This paper introduced a potentially promising framework for predictive maintenance in the field of machine milling. The framework is based on a modified firefly algorithm, and it was tested and verified on two publicly available datasets. The aim of the framework is to monitor the wear of the milling tools and signal when replacement is required, thereby avoiding tool breakdown, saving time, and thus improving the productivity of the machine. Due to the intensive nature of optimization and data processing tasks, which involve large-scale multidimensional datasets and iterative model refinement, the proposed method benefits from high-performance computing resources. The algorithm’s design supports parallel and distributed execution, enabling real-time processing of sensor data and rapid decision-making required for autonomous manufacturing systems.