Machine learning for predictive maintenance in self-healing software services

Nagaraj Bhadurgatte Revanasiddappa *

Individual Researcher, Engineering Technology Leader, USA.
 
Review
International Journal of Science and Research Archive, 2022, 05(01), 162-181.
Article DOI: 10.30574/ijsra.2022.5.1.0027
Publication history: 
Received on 17 December 2021; revised on 25 January 2022; accepted on 28 January 2022
 
Abstract: 
Automated self-repairing systems, catalyzed by predictive maintenance and ML, are the new formative model in today’s software industry. Predictive maintenance is used widely for anticipating and avoiding system failures, and it forms a critical element in improving reliability and performance of software services. This article focuses on machine learning and application like predict and prevent maintenance and self-healing system that helps minimize downtimes, increases overall system performance of a system and selects optimal use of resources. The first part gives an overview of the development of software systems whereby an analysis was done on the general transition from simple static systems to autonomous systems that can self repair. It talks about predictability maintenance as a way of foreseeing failures, therefore avoiding undue disasters which are likely to inconvenience end users. The article then proceeds to explain the underlying ideas of predictive maintenance and differentiating it from the preventive and reactive kinds. This paper elaborates the methodology of how predictive maintenance integrates with self-healing and also with focus on without human intervention techniques used in autonomous systems wherein faults are corrected in real time. The discussion goes on to Machine learning methods ubiquitous in the predictive maintenance. Fault detection is tackled in supervised learning, and anomalous patterns in system logs are detected in unsupervised learning; reinforcement learning is applied to form recovery models to enhance self-healing mechanisms. The article goes on to discuss some of the difficulties in data quality, scaling of models and the combination of ML systems into existing structures, with solutions put forward including the use of cloud and the use of mix-model solutions. The use of predictive maintenance in the real world includes; In the healthcare sector and in cloud computing to name but a few. These case studies illustrate how organizations deploy these technologies for dependable and high performing operation in critical applications. Last but not least, the article brings vision for the future, with new interesting enhancements in self-healing solutions and the progression of AI/ML making the shift towards autonomous and self-optimizing systems.
 
Keywords: 
Self-healing software; Predictive maintenance; Machine learning; Fault detection; Anomaly detection
 
Full text article in PDF: 
 
Copyright information: 

Copyright © 2022 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution Liscense 4.0