MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202621071914 A) filed by Vit Bhopal University on June 10, 2026, for System And Method For Trust-Aware Explainable Artificial Intelligence In Safety-Critical Environments.
Inventors include Kolhatkar, Tanishq; and Bose, Subash Chandra.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: A trust-aware explainable artificial intelligence system (100) for safety-controlled decision-making in a safety-critical environment may include one or more data sources (102) and a processing system (106) having one or more processors (106A) and a memory (106B) storing executable instructions. The processing system (106) may receive input data from data sources (102), generate an output via a prediction module (106D-4), and compute a composite TrustScore via a trust evaluation module (106D-5) by aggregating reliability parameters including model confidence, model uncertainty, input data quality, and prior performance indicators. An explainability module (106D-6) generates human-interpretable explanations. A decision gate mechanism (106D-7) performs trust-based evaluation by comparing the TrustScore with a threshold to classify the output. A human interaction interface (108) presents the output, TrustScore, and explanations for validation or override. The processing system (106) then produces a control signal and initiates a hardware-based safety action based on the classification or human input.
Disclaimer: Curated by HT Syndication.
