MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641094180 A) filed by Akhil Aneel Deshpande; Dr C Lalitha; Dr Julie Patel; Tata Sowjanya Kumari; Dr Kalpesh V. Gohil; Shabana Sultana; J. Mary Angeline; and Rajendra Singh Rajput on August 04, 2026, for Machine Learning-Based Predictive Maintenance System For Industrial Iot.
Inventors include Akhil Aneel Deshpande; Dr C Lalitha; Dr Julie Patel; Tata Sowjanya Kumari; Dr Kalpesh V. Gohil; Shabana Sultana; J. Mary Angeline; and Rajendra Singh Rajput.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: Predictive maintenance (PdM) has emerged as an essential approach for enhancing the efficacy and dependability of industrial equipment. Incorporating machine learning techniques into a predictive maintenance framework offers a viable approach for enhancing maintenance plans, averting equipment malfunctions on the production line, and minimising downtime. This study introduces a data-centric methodology for identifying malfunctions in industrial machinery through sensor information. The approach seeks to enhance system efficiency, leading to financial savings such as reduced energy usage and maintenance expenses. The methodology presented in this study involves the creation of a Predictive Maintenance system tailored for yarn production machinery utilising machine learning techniques. The efficacy of predictive maintenance applications relies on the judicious choice of machine learning techniques. The Industrial Internet of Things (IIoT) facilitates the collaboration of machines, individuals, cloud computing, and analytics, hence enhancing the efficiency and output of industrial operations. Historically, the upkeep of industrial machinery employed techniques like periodic maintenance (PM) or condition-based maintenance (CBM). Preventive maintenance frequently results in the squandering of human resources and materials, as maintenance is often superfluous and may be deferred. CBM necessitates an elevated degree of specialised expertise to establish the threshold values. Unlike these methods, predictive maintenance represents a more efficient and effective strategy that entails observing the condition and health of industrial machinery to detect potential failures and risks prior to their escalation into incidents resulting in significant property damage, production disruption, or loss of life. Forecasting necessitates machine learning algorithms that rely on extensive datasets for every system element. This article presents a review of the literature about industrial IoT and the implementation of deep learning and artificial neural networks to produce rapid forecasts for maintenance and various IoT applications. A technique for the upkeep of machinery and equipment known as predictive maintenance (Pd.M.) aims to foresee potential issues and prevent them from occurring. The aim is to extend service life by precisely forecasting the timing of faults and averting them. Projections rely on present circumstances and amassed expertise.
Disclaimer: Curated by HT Syndication.
