Humanity is about to be handed almost unimaginable power, and it is deeply unclear whether our social, political, and technological systems possess the maturity to wield it.
Dario Amodei, CEO and co-founder of Anthropic; widely recognized as a leading voice in AI safety
In this article, I’ll share the specific topics that will be addressed in the FMEA and AI series.
If you haven’t already done so, I suggest reading my previous article, New Series: FMEA and AI. It discusses my high-level view of the use of AI in support of FMEA.
I’ll repeat an excerpt for emphasis:
AI has the potential to serve as a powerful accelerator and idea generator in FMEA, as well as ability to scan and integrate volumes of data, which can enhance FMEA efficiency and breadth. However, its outputs demand rigorous human oversight due to potential risks of inaccuracy, bias, lack of true understanding, and inability to handle novelty or ethicalconsiderations inherent to safety-critical engineering. FMEA teams provide human insight, critical thinking, synergy, creativity, and innovation that is essential to achieving effective FMEAs, both for current FMEA applications and future updates. When using AI to support FMEAs within a company, it is essential that AI is trained on the correct fundamentals of FMEA.
Topics to be included in this series
Here is an outline of the topics for articles in this series. Each topic will have one of more articles, and will be explored for how AI can support FMEA, the cautions to be aware of, and what level of human involvement is essential for success.
Topic 1: How can AI support FMEA planning?
When engaging in FMEA planning, consider this quote from Benjamin Franklin: “If you fail to plan, you are planning to fail.” The message: always begin by prioritizing the set of FMEAs that need to be done in support of a given project, including the types of FMEA, when each will be done, and who will lead the FMEA team.
In this topic, I will share how AI can support the development of a workable FMEA plan, including the cautions.
Topic 2: How can AI support FMEA preparation?
I’ll begin here with a quote from Confucius: “Success depends upon previous preparation, and without such preparation there is sure to be failure.” The message: FMEA will not be successful without thorough and proper preparation steps.
There are five key steps to FMEA preparation. There can be other preparation steps, depending on company or organization FMEA policy, but these five are essential for all applications.
- Determine FMEA scope
- Make FMEA scope visible
- Establish ground rules & assumptions
- Gather information
- Assemble the correct FMEA team
Each of these preparation steps will be explored for how AI can best be used to support them, both the positives and the cautions.
Topic 3: How can AI support identification of FMEA elements?
FMEA procedure includes certain data elements, which are identified as part of the technical risk analysis. These typically include: Item, Function(s), Failure Mode(s), Effect(s), Cause(s), Special Characteristic(s), Current Controls, and Compensating Provisions. There may be others in your procedure.
Each of the FMEA data elements will be explored for how AI can support, including cautions.
Topic 4: How can AI support assessing and prioritizing risk for each FMEA?
Prioritizing risk in an FMEA includes assessing the severity, occurrence, and detection levels for failure modes, effects, and causes. I’ll take up each of these separately, and show how AI can support, and what to be concerned about. I’ll also discuss how AI can support the assignment of priority for corrective actions.
Topic 5: How can AI support reducing risk to acceptable level for each FMEA?
Every FMEA should be continued until risk is reduced to an acceptable level. This involves recommending corrective actions, getting them executed, and reassessing risk. AI can be very helpful in these actions, but with certain cautions, which I will explain. I will share how AI can be helpful in recommending corrective actions and tracking them to completion.
Topic 6: How can AI support certain follow-up steps for each FMEA?
There are certain follow-up steps that help to maximize the value of FMEA, such as linkages to other methods, change management, version control, FMEA updates, and quality objectives. Each of these follow-up steps can be supported by AI. How to do this and what cautions to be concerned about will are explored.
Topic 7: What are the essential human roles in FMEA that cannot be done by AI?
Critical to successfully implementing AI support for FMEA is to understand where human involvement is essential. What are the essential roles that humans play in creating FMEAs that are support by AI? This topic will be explored in depth.
Topic 8: Why is a deep understanding of FMEA principles necessary for successful AI support?
Key FMEA principles transcend the rich variety of FMEA applications and underpin the success of AI integration. FMEA principles are covered in my books and articles. No proem of AI support for FMEAs will work without being informed by correct FMEA principles.
Topic 9: What is management’s role in integrating AI with FMEA?
Management has an important role in the implementation of effective FMEAs in any company of organization. This topic will explore the specific roles of management in successfully implmentting FMEA, and in achieving maximum AI support.
Join me each month as we explore these topics.
Leave a Reply