MUMBAI, India, July 30 -- Intellectual Property India has published a patent application (202641088591 A) filed by Mr. Aman Tomar; Dr. Ashok Kumar M; Mr. Anish Antony; Ms. Sagarika Kar Chowdhury; Mr. Kadam Rushikesh Raju; Mr. Yogananth S V; and Ms. Sudhiksha A on July 21, 2026, for Explainable-Ai Audit Trail System And Method For Automated Quality Rejection Decisions In Manufacturing Inspection.
Inventors include Mr. Aman Tomar; Dr. Ashok Kumar M; Mr. Anish Antony; Ms. Sagarika Kar Chowdhury; Mr. Kadam Rushikesh Raju; Mr. Yogananth S V; and Ms. Sudhiksha A.
The application for the patent was published on July 24, 2026, under issue no. 30/2026.
Abstract: Explainable-AI audit trail system and method for automated quality rejection decisions in manufacturing inspection Abstract: A system and method automatically generate a structured, traceable explanation record for automated accept/reject decisions made by a multi-sensor AI/ML quality-inspection system on a manufacturing line. A synchronized multi-sensor capture rig timestamps and hardware-triggers two or more sensor modalities (e.g., visual, thermal, acoustic) from a common event source, including for legacy equipment lacking shared network time. Upon a reject decision, an attribution pipeline computes a per-modality contribution score, a spatial attribution map localizing the contributing region(s) of the part, and a nearest-neighbor comparison against a previously accepted part with a computed difference score. A record generator automatically assembles this data into a standardized decision audit record mapped to a quality-traceability schema, without human authoring, and may cryptographically bind the record to the underlying synchronized sensor data for audit integrity.
Disclaimer: Curated by HT Syndication.
