Assembly Quality Management using system intelligence (Research)
Process Failure Mode and Effects Analysis (pFMEA) is an essential tool to detect and address potential failures rooted in the assembly process, linked to 4M causes (Man, Machine, Material, environMent). As a team-oriented procedure, it is well established in different industries and even mandatory in sectors like automotive. Current pFMEA however is expensive, subjective and slow and the analysis too narrow to capture propagated effects beyond single-point failures. The analysis is difficult to update after mitigation actions and it is hard to capture the learnings for the design of next generations of the assembly system. AQUME will develop a data-supported Process Failure Modes and Effects Analysis and Actioning (pFMEA2) decision support system, to better activate quality assurance with process and quality data as well as expert knowledge. A reasoning engine built around Bayesian networks and backed up by a linked process and quality knowledge graph architecture will make pFMEA more objective, automated and repeatable. Dedicated interfaces will activate targeted information gathering from operators during production. AQUME moves the current static pFMEA towards an online, continuously updated pFMEA2 ("Analysis & Actioning") that is able to automatically trigger MES actions to solve issues like tool replacement. As a continuous process, it will be able to monitor changes in the assembly system, reassess previous analysis and better capture and document the learnings of previous versions of the system towards new designs with thereto-developed interactive visualizations and interfaces.
Period of project
01 January 2023 - 31 December 2026