MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202611104898 A) filed by Panipat Institute Of Engineering And Technology Piet on August 31, 2026, for Graph-Based Multi-Agent System And Method For Proactive Substitute Node Allocation And Interruption-Free Distributed Learning.
Inventors include Mr. Anurag Vashist; and Mr. Piyush Babele.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: The present invention provides a graph-based multi-agent system and method for proactive substitute node allocation and interruption-free distributed learning. A distributed computing infrastructure comprising a plurality of participating computing nodes is represented as a dynamically updated graph, and intelligent software agents are associated with respective nodes. Operational parameters comprising node health, computational resources, communication characteristics, workload, connectivity, availability and trust information are monitored during distributed learning. A predictive failure analysis engine determines an interruption probability for a participating node. When the probability satisfies a failure criterion, candidate substitute nodes are identified from the graph and evaluated according to resource availability, graph connectivity and operational suitability. A selected substitute node is allocated before interruption of the vulnerable node, and learning state and computational responsibility are proactively migrated thereto. The graph topology is subsequently updated and distributed learning continues through the resulting computing configuration, thereby reducing synchronization delay, computational loss and post-failure recovery overhead. Accompanied Drawing [FIG. 1]
Disclaimer: Curated by HT Syndication.
