User Story: AI-Powered Automated Welding Quality Inspection ID: TMBS-WELD-01 Title: Real-time Welding Defect Identification Priority: High User Story Statement: As a Quality Control Engineer at TMBS, I want the AI inspection system to automatically analyze live video feeds from the welding stations, So that I can identify and categorize weld defects instantly and ensure only high-quality modules proceed to the next assembly stage. Acceptance Criteria (AC): AC 1: The system must accurately classify the welding output into the following categories: burn_through , crack , porosity , undercut , overlap , or good_weld . AC 2: The system must provide real-time inference during the welding process, triggering an alert if a defect is detected. AC 3: If a defect (0 through 4) is identified, the system must log the specific defect type and timestamp to the central tracking database. AC 4: The inspection UI must clearly highlight the detected defect on the video feed to assist the...