MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641096876 A) filed by Vellore Institute Of Technology on August 11, 2026, for Autonomous Cognitive Infrastructure Operating System For Self-Evolving Runtime Management.
Inventors include Jagadeesh; and M S Shabreena.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: WE CLAIM: 1. An Autonomous Cognitive Infrastructure Operating System (ACIOS) (100) for self-evolving runtime management of cyber-physical infrastructure, comprising: an Infrastructure Data Acquisition Layer (102) configured to acquire heterogeneous operational data from distributed cyber-physical infrastructure sources and normalize the data into standardized runtime infrastructure events; a Runtime Infrastructure Genome Engine (RIGE) (104) configured to generate and maintain a Runtime Infrastructure Genome (RIG) for each infrastructure entity, wherein each Runtime Infrastructure Genome comprises an executable runtime object with continuously evolving computational state vectors; a Dynamic Dependency Fabric Engine (DDFE) (106) configured to construct executable dependency objects among infrastructure entities, wherein each dependency object comprises adaptive execution parameters; a plurality of Specialized Autonomous Infrastructure Agents (108) configured to analyze runtime infrastructure state and generate structured Intent Objects; an Intent Convergence Execution Engine (ICEE) (110) configured to synthesize the Intent Objects into a unified executable infrastructure action through dependency evaluation, conflict resolution, and adaptive prioritization; a Predictive Twin Materialization Engine (PTME) (112) configured to selectively materialize execution-relevant digital twin regions based on predicted operational requirements; an Autonomous Decision Verification Engine (ADVE) (114) configured to validate the unified executable infrastructure action against multi-dimensional constraints prior to execution; an Infrastructure Execution Fragment Repository (IEFR) (118) configured to store reusable Infrastructure Execution Fragments decomposed from executed infrastructure actions; and an Infrastructure Evolution Kernel (IEK) (120) configured to continuously refine the Runtime Infrastructure Genome, dependency relationships, and execution policies using operational feedback, thereby establishing a closed-loop autonomous infrastructure computation framework. 2. The system of claim 1, wherein each Runtime Infrastructure Genome comprises continuously evolving computational vectors including a Physical Vector, a Capability Vector, a Resource Vector, a Constraint Vector, a Dependency Vector, an Operational Vector, a Recovery Vector, a Trust Vector, a Mission Vector, and an Evolution Vector. 3. The system of claim 1, wherein the Dynamic Dependency Fabric Engine (106) is configured to represent dependency categories including spatial dependencies, operational dependencies, communication dependencies, maintenance dependencies, energy dependencies, mobility dependencies, temporal dependencies, safety dependencies, and resource dependencies. 4. The system of claim 1, wherein the adaptive execution parameters of each dependency object comprise a dependency type, a dependency weight, an execution priority, a propagation coefficient, a temporal persistence value, a confidence level, and a risk influence value. 5. The system of claim 1, wherein the Specialized Autonomous Infrastructure Agents (108) comprise an Energy Agent, an HVAC Agent, a Water Agent, a Security Agent, a Maintenance Agent, a Mobility Agent, a Sustainability Agent, an Emergency Agent, and an Occupancy Agent. 6. The system of claim 1, wherein the Intent Convergence Execution Engine (110) is configured to perform intent collection, intent classification, constraint analysis, dependency evaluation, conflict detection, recursive intent convergence, adaptive prioritization, execution optimization, and global intent synthesis. 7. The system of claim 1, wherein the Predictive Twin Materialization Engine (112) is configured to perform infrastructure state prediction, execution region identification, incremental twin materialization, runtime state synchronization, and predictive infrastructure simulation. 8. The system of claim 1, wherein the Autonomous Decision Verification Engine (114) is configured to validate each execution plan against safety constraints, regulatory compliance requirements, organizational policies, cybersecurity policies, resource availability, infrastructure constraints, accessibility requirements, sustainability objectives, and operational continuity requirements. 9. The system of claim 1, wherein each Infrastructure Execution Fragment comprises a runtime infrastructure state, a trigger event, an intent structure, a dependency snapshot, an execution sequence, resource utilization data, an operational outcome, a recovery strategy, performance metrics, and a confidence level. 10. The system of claim 1, wherein the Infrastructure Evolution Kernel (120) is configured to refine Runtime Infrastructure Genomes, dependency relationships, execution priorities, scheduling policies, optimization strategies, agent decision parameters, intent generation parameters, and Infrastructure Execution Fragments.
Disclaimer: Curated by HT Syndication.
