MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641091782 A) filed by Dr Sajeeda Parveen Shaik.; Dr. G. Yuvaraj; Mrs. Rajasugirtha N; Abhishek Ramesh Jaware; and Ajit Shivaji Dange on July 29, 2026, for A System And Method For Autonomous Self-Healing Computing Infrastructure And Intelligent System Recovery Using Artificial Intelligence.
Inventors include Dr Sajeeda Parveen Shaik.; Dr. G. Yuvaraj; Mrs. Rajasugirtha N; Abhishek Ramesh Jaware; and Ajit Shivaji Dange.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: ABSTRACT A System and Method for Autonomous Self-Healing Computing Infrastructure and Intelligent System Recovery Using Artificial Intelligence The present disclosure relates to a system and method for causally verifying faults and automatically recovering distributed computing infrastructure using Artificial Intelligence. The system collects time synchronized telemetry from computing/storage/network and messaging resources, and builds a dependency topology. A causal inference engine identifies candidate causes of faults and selects a reversible diagnostic perturbation within a service-impact limit and evaluates an infrastructure response resulting from the perturbation to verify an actual fault. The resources affected are identified by a fault-boundary compiler, and the corresponding checkpoint, transaction, network-session and message states are captured by a recovery-capsule generator. A continuity-token generator cryptographically binds a compute checkpoint epoch, a storage journal watermark, a network-flow epoch, and a message-queue offset. A recovery coordinator stages replacement resources and enables activation upon verification of a continuity-token. After activation, hardware and service invariants are evaluated to decide whether to commit the recovered configuration or revert to a previous configuration, reducing incorrect remediation, recovery bandwidth, cross-resource inconsistency, cascading failures, and service interruption across distributed infrastructure operations.
Disclaimer: Curated by HT Syndication.
