MUMBAI, India, July 7 -- Intellectual Property India has published a patent application (202641076654 A) filed by Mrs. K. Bavani; Ms. N. Subasri; Ms. Karthiga. R; Dr. Arivu Selvam B; Ms. L. Saranya; Ms. Malarkodi M; and Mr. S. Palpandi on June 20, 2026, for Machine Learning-Enabled Cloud Platform For Autonomous Iiot Operations.
Inventors include Mrs. K. Bavani; Ms. N. Subasri; Ms. Karthiga. R; Dr. Arivu Selvam B; Ms. L. Saranya; Ms. Malarkodi M; and Mr. S. Palpandi.
The application for the patent was published on July 03, 2026, under issue no. 27/2026.
Abstract: Machine Learning-Enabled Cloud Platform for Autonomous IIoT Operations Abstract A system using machine learning within a cloud setup supports self-managed tasks in industrial IoT settings. Unlike older methods, which rely heavily on human oversight and adjustments, this approach reduces inefficiencies tied to resource usage. When workloads shift or machinery differs across sites, past systems often fail to spot problems quickly. Response delays lead to higher costs and uneven performance outcomes. Without constant monitoring, traditional models struggle to maintain consistent service standards. Despite its flaws, the system works by linking three main parts. First comes secure collection of varied data types - gathered continuously from scattered sensors, machines, and control units. What drives insight is a smart analysis module: it applies labeled-data classification to spot operating conditions, uses grouping methods to flag anomalies, while also drawing on trial-and-error learning, mixing near-policy updates with deep Q strategies to shape independent decisions over time. Then follows an active response mechanism adjusting task assignments dynamically, balancing workloads between nearby edge devices and remote servers, triggering early repairs before breakdowns occur, plus tweaking production steps with precision. Communication flows both ways - results after each move are recorded, feeding back into the loop so models refine themselves gradually, adapting without needing constant oversight. Because machine learning classifiers detect unusual activity linked to cyber threats or system failures early, security strengthens alongside dependable operations. Encryption standards apply widely, while communication systems continue functioning even if parts fail. Scalability adjusts smoothly - this design fits extensive networks across many industrial locations. Architecture holds steady under shifting demands, maintaining performance without disruption. Because of how the system works in practice, key areas show clear progress - fewer surprise shutdowns happen now. Efficiency in using power and materials improves when operations run longer without help. After sudden shifts, balance returns more quickly than before. Performance stays closer to required standards over time. This approach creates a full cycle of automatic adjustments, helping machines operate independently with greater reliability. Human experts need to step in less often as stability grows under changing conditions. This version appears in smooth, formal blocks designed for immediate inclusion within a patent document. Formatted neatly into eleven- or twelve-point typefaces like Times New Roman or Arial, it fills just one A4 sheet - given regular margins and even edges along both sides. Not a single acronym shows up in the main section, following earlier guidance. Still, clarity holds steady without shortcuts or condensed forms muddying meaning.
Disclaimer: Curated by HT Syndication.
