MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641125330 A) filed by Vellore Institute Of Technology on October 21, 2026, for Adaptive Trust-Driven Forensic Intelligence For Autonomous Crime Scene Reconstruction And Explainable Evidence Correlation Using Iot And Spatial Ai.
Inventors include Dr. Gunasekaran G; Praveen Kumar V; Naveen Kumar K; Harish Y; and Suresh Krishna K.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: ABSTRACT The present invention relates to an adaptive trust-driven, multi-modal forensic intelligence system and method for autonomous crime scene reconstruction, probabilistic evidence correlation, and explainable investigation. The system comprises an integrity-screened sensor trust engine, adaptive evidence fusion engine, temporal evidence decay-aware confidence engine, probabilistic knowledge graph engine, confidence-adaptive spatial reconstruction engine, human-in-the-loop recalibration engine, federated trust-transfer engine, and cryptographic reasoning lineage manager. The system detects anomalous or compromised sensor observations, recursively updates sensor trust coefficients, and performs trust-weighted evidence fusion. Environmental telemetry and elapsed time determine category-specific evidence decay factors for recalibrating evidence confidence and collection priority. A probabilistic knowledge graph supports evidence relationship updates and counterfactual reasoning, while three-dimensional reconstruction associates spatial objects with continuously updated confidence values. Investigator feedback enables bounded trust recalibration, and federated aggregation facilitates cross-case trust transfer without transmitting raw evidentiary data. A hash-chained ledger preserves verifiable investigative processing records, thereby improving forensic reliability, explainability, traceability, and auditability.
Disclaimer: Curated by HT Syndication.
