Reliability Engineering and Artificial Intelligence
Abstract
Carl and Chris discuss the changing opportunities and challenges with AI and reliability engineering. What are the positive interactions, and where should we be concerned?
Key Points
Join Carl and Chris as they discuss their views on the roles that AI can play in reliability programs. Topics include:
- What are the limitations of AI in supporting reliability programs?
- One concern is if AI is being oversold, and replaces necessary human involvement.
- Another concern is the possibility of AI moving a company towards mediocrity.
- AI can be useful to help ensure you don’t miss something critical.
- AI cannot perform an FMEA. It can augment an FMEA. It can provide excellent input to FMEA.
- Discussion around having an AI bot as part of an FMEA team. However, it may lessen creativity.
- Humans cannot see what is missing. AI generated input can retard human innovation and creativity. For this reason, it may help to begin with human creativity before bringing AI generated input.
- The purpose of FMEA is not to fill out a form. It involves surfacing risk, and reducing risk to an acceptable level. FMEA helps to improve the design or manufacturing process.
- Example: ask AI if MTBF is a good reliability metric. The answer is not very useful.
- Tip: review the citations listed with AI responses.
- Tip: Use AI to fill in the gaps, not the other way around.
Enjoy an episode of Speaking of Reliability. Where you can join friends as they discuss reliability topics. Join us as we discuss topics ranging from design for reliability techniques to field data analysis approaches.
SOR 1095 Reliability Engineering and Artificial IntelligenceCarl S. Carlson
- Social:
- Link:
- Embed:
Show Notes
What a great podcast. With any tech, we should be looking to use it to boost what we have in place and drilling down more to capture those points of OEE that create transactional ROI. We should not be expecting some ‘hack’ to high reliability metrics. It’s doesn’t work in any category. If there is no suitable risk management policy in place then getting an unknown rogue (AI) to tell you where your risk lies is asinine and and idea that comes from people that are afraid to do the work.
I do however feel that the organizations that refuse to use AI will suffer to some degree which will be cited along the same string of excuses as to why they don’t use a CMMS effectively. That group however will be smaller than those businesses, projects, etc that use AI incorrectly. As Chris makes mention a couple times, use it to identify gaps then go to work. At best for me at this point, AI speeds up my Google searching and edits my poor grammar, which apparently is getting better because I didn’t use it for this comment. Thanks for the share guys. Cheers.