Sustainable Manufacturing through Predictive Maintenance: A Hybrid Jaya Algorithm and Sea Lion Optimization and RNN Model for Industry 4.0

• Main Authors: Nimma, Divya, Silambarasan, R, R, Venkateshkumar, Manchanda, Mahesh, Changala, Ravindra, Kiran Bala, B
• Format: Conference Proceeding
• Language: English
• Subjects: Accuracy; Convolutional Neural Networks; Hybrid Optimization; Industries; Industry 4.0; Marine animals; Optimization; Prediction algorithms; Predictive maintenance; Production; Protection; Recurrent neural networks; Sustainable development
• ISSN: 2768-0673
• DOI: 10.1109/I-SMAC61858.2024.10714701

Sustainability in production has grown to be more and more essential as industries aim to reduce their environmental impact at the same time as boosting productiveness. Predictive protection offers a proactive approach to foresee gadget failures before they occur, optimizing resource use and supporting sustainability goals. This study offers an advanced predictive upkeep version that integrates a hybrid of the Jaya algorithm and Sea Lion Optimization (SLO) with Recurrent Neural Networks (RNNs) inside the industry 4.0 framework. The hybrid Jaya-SLO approach combines the simplicity and effectiveness of the Jaya set of rules with the robust seek abilities of SLO to optimize preservation scheduling and minimize system breakdowns. The addition of RNNs complements the version's predictive electricity via studying historical time-series records of device overall performance, leading to noticeably accurate failure forecasts. Implemented in Python, this version achieves a topnotch accuracy of 98%, surpassing traditional predictive renovation methods. This research contributes to the existing literature by employing a novel combination of hybrid Jaya RNNs and Sea Lion Optimization for the predictive maintenance for Industry 4. 0. Convolutional Neural Networks (CNNs) are utilized for function extraction, uncovering complex styles within the statistics that further refine the predictive version. The model's overall performance is evaluated the usage of metrics together with accuracy, precision, and recall, and is integrated into current maintenance frameworks for realistic deployment.