AI in quality management

The AI can be used for FMEA in two targeted cases, when creating a new FMEA and when updating an existing one, to speed up the work and ensure quality. Let's get started. In the first case, a new FMEA is created, and the AI does the groundwork so the team reaches a complete result faster. When the project reaches a milestone, the AI clarifies in the chat what is new and what remains existing technology, and proposes the core team to match the discipline. It then requests input from the departments with a deadline and checks whether everything is complete. From old FMEAs, foundation FMEAs, lessons learned and 8D reports, it retrieves comparable knowledge, which is thus also available to inexperienced teams. From this, it builds the structure tree with process steps and 4M, proposes functions and asks where information is missing. For every function, it proposes failure effects, failure modes and causes with their source, so every proposal remains traceable. For severity, occurrence and detection, it proposes values with justification and calculates the action priority, and the team makes the decision. In the review, it assigns review tasks by expertise and presents interfaces to both disciplines, so the final meeting only clarifies open points. Finally, it derives the test planning, transfers the FMEA to your FMEA tool and marks special characteristics for the control plan and DFMEA. In the second case, an existing FMEA is updated, and the AI makes sure changes are carried over quickly and completely. If the trigger is a closed 8D report, the AI reads the confirmed root cause from D4 and the actions from D5 to D7. It then checks what is affected, down to linked and similar FMEAs, and clarifies whether the cause was rated incorrectly or was missing entirely. The change proposal only affects the specific places, for example a higher occurrence rating and new actions from the 8D, and everything else remains unchanged. Only the experts whose area is affected need to review, and often an approval without a meeting is enough. After approval, the FMEA is versioned with a reference to the 8D report, the AI shows the follow-up documents to update, and D7 can be closed.

This video shows how the AI creates inspection plans directly from the drawing and detects deviations in your measurement data before parts are out of tolerance. Let's get started. For a new inspection plan, the AI reads the drawing, identifies dimensions, tolerances and geometric tolerances and numbers the characteristics directly in the drawing. For every characteristic, it proposes inspection method, inspection equipment and inspection frequency and marks special characteristics from the FMEA and customer requirements. Your team reviews the draft and approves it, and the inspection plan is then transferred to your CAQ system. During operation, the AI automatically takes over the measurement files from your measuring machines and shows by plant and part which characteristics are out of tolerance. If a characteristic is heading towards the tolerance limit, the AI detects the trend and warns before the limit is reached. You confirm each alert or set it to monitor, and you document permanent deviations as a concession with justification and due date.