What are the best tools and methods for FMEA data collection and analysis?

As with any analysis, it must have clear boundaries (i.e., scoped). For example, if this is a Process FMEA (PFMEA), are you including the product "movement" into and out of the process or is this being separated into a Logistics LFMEA? Also important is defining who the "customers" of the process are. Is this process feeding a higher tier supplier or is the product being delivered directly to an OEM?

Ferramentas como diagramas de pareto e matriz de criticidade podem ajudar na decis?o de por qual ativo come?ar a análise. Normalmente ativos com grandes custos de manuten??o, ativos com grandes quantidades de parada de processo e ativos, cuja falha podem levar a grandes perdas materiais e vidas. Estes s?o os ativos que normalmente devem ser priorizados.

For DFMEA, a functional analysis (incl. defining product characteristics) is required across the product BOM. Each BOM entity requires clear functional descriptions of the activity(s) they perform. They are connected to form a function network (model) defining the product function...by design. The team also captures all relevant functional requirements to clarify functional verification with testing. The resulting wholistic function model represents the product in operation. Next, failures can be effectively defined & connected in cause-effect models that become the basis for the FMEA failure effect-mode-cause relationships. This analysis requires dedicated software to perform well except for the most basic of products or processes.

Em minha experiência, a base de dados de interven??es de manuten??o do sistema ERP da empresa é o melhor caminho. Mesmo que necessite de saneamento e diversas filtragens, antes de poder fazer inferências sobre as informa??es contidas na base. A base de dados da empresa é a melhor representa??o para apontar os modos de falha específicos do processo produtivo da propria empresa. Base de dados genérica podem n?o representar a criticidade real de algum modo de falha específico do item que está analisando. Porém, base de dados genéricas s?o uma segunda alternativa como fonte de dados para análise de processos em empresas que ainda n?o possuem um banco de dados consolidado.

There are new methods in data collection leveraging artificial intelligence in failure analysis and using it to search for diverse sources online, gathering them, and referencing their primary sources. Obtaining data has become easier with the proper use of technology.

When applying FMEA to maintenance, automation of data collection and integration is paramount. This involves automating the collection of failure history, maintenance logs, and machine data, and integrating them into a centralized system. By automating these processes, organizations can streamline data gathering, improve accuracy, and enable real-time analysis. This facilitates proactive identification of potential failure modes and enhances the effectiveness of preventive maintenance strategies, ultimately optimizing equipment reliability and performance.

When no actual test data exists that verifies the actual detection performance a detection approach for a specific failure mode then it should be scored as if there is 0 probability of detection. This will force action to test the detection effectiveness of high risk failure modes. I have worked on too many problems where wishful thinking was at large and detection was subjectively scored at 100% effective without proof, and then a quick physical test of the detection approach revealed a much lower level of detection.

With the function and failure models defined, the failure causes are assessed for existing preventative and detection actions. These are subjectively scored according to their effectiveness in preventing the cause, and detecting the failure mode or cause. Adding the severity of the effects to this mix, various scoring metrics can be used (e.g., AP, RPN, Risk Matrix, etc.) to prioritize the risks.

The contemporary use of data analysis tools is crucial for researchers to comprehend and evaluate the root causes of failures across various domains. Delving deep into available data provides comprehensive insights into factors contributing to failure, whether in products, processes, or human behaviors. This analysis aids in making crucial decisions and developing procedures to mitigate future failures. By employing data analysis, organizations grasp failure causes and take necessary actions to enhance performance and reduce future risks. Data analysis isn't just for analyzing past occurrences; it's a powerful tool for predicting the future and making strategic decisions based on strong foundations and in-depth knowledge.

Em minha experiência, na maioria dos casos estudados, a??es simples como adi??o de um passo de inspe??o ou preventiva no plano de manuten??o é suficiente para mitigar o modo de falha. A própria equipe de chao de fabrica, que atua diretamente com o ativo tem os melhores insights na busca da solu??o. Cabe ao coordenador do estudo saber direcionar em a??es eficazes e duradouras.

FMEA is a continuous process that requires regular review and updates over time. Therefore, having an automated tool is crucial to facilitate this task quickly and efficiently, preventing the program from becoming a burden or falling into disuse. Automated tools streamline the FMEA process, enabling organizations to keep pace with changes, identify emerging risks, and implement timely corrective actions, thus ensuring the effectiveness of their maintenance strategies while minimizing operational disruptions

I agree, but the path to efficient analysis lies in selecting software specifically designed to analyze such data effectively. Choosing a tool tailored to analyze failure data must be customized for each sector of business: startups, small and medium-sized enterprises, large corporations, etc.