Reliability Basics and AI with Fred Schenkelberg
Artificial Intelligence has become a hot topic today with organisations attempting to find ways of applying it. The fact is, however, that a good number haven’t found a compelling case for the adaption. Fred of Accendo Reliability joins us on this episode to break down what AI means for the reliability industry.
Key highlights from this interaction are:
- How reliability is influenced by AI
- Why people are jumping on the ‘trend’ without looking at the relevance
- Considerations to make before adopting AI for reliability
Today, more assets are becoming smarter and in the process generating large quantities of data. Ironically, even with countless data points flowing into databases, there is yet to be sufficient information that generates value.
With regards to AI, tools for performing analytics are a step in harnessing insights. However, the underlying logic of AI algorithms comes from having a good understanding of the systems as well as the basic principles of mathematics. Just because the competition is embracing AI does not mean you should also buy the next plug-and-play solution.
How is reliability influenced by AI?
In the competitive world today, maintenance and reliability teams are discovering ways in which AI can help them:
- Optimize maintenance schedules and predict abnormal performance
- Select appropriate conditions for equipment performance
- Increase process efficiency
The technology revolution is only just building momentum, and those who embrace it dynamically are the ones poised to lead their industries.
Why some people without AI experience are still jumping in?
Because of its seemingly trendy nature, AI has attracted many teams to start tinkering with it. The industry is facing something similar to the ‘shiny object syndrome’
A fraction of those who try implementing AI strategies without a fundamental understanding of the need end up with a model that produces inconsistent results. The wrong approach is staring off with an ‘AI solution’ and looking for the problem. It is therefore critical for personnel in charge of reliability to have a good understanding of their existing operational difficulties, before figuring out where AI can plug into it.
Key considerations to make before diving into AI for reliability
Having the fundamentals down means that from a design point, both intrinsic and extrinsic influences get factored into the performance consideration. When going through designing an AI solution, you can then take a procedural getting ideas like:
- Identify the existing problem
- Interrogate ways in which AI can improve it in comparison to the existing state
- Consider the influence of intrinsic and extrinsic factors like seasonality, product mix etc, and how they would affect the AI model
- Gather sufficient historical data that will be used to train the model. Ensure it covers as many performance scenarios as possible
- Once trained with historical data, try testing the AI model using appropriate methods e.g supervised learning, unsupervised learning etc
Other important considerations are that you should have a skilled data scientist or statistician working with domain experts, so as to adapt these algorithms to your facilities day-to-day scenarios.
Understanding the problem is a key component for successful deployment of AI. You need to ask yourself:
- What problem are we trying to solve?
- Why is the existing infrastructure not fast enough or powerful enough, to require AI?
After this problem definition phase, you need to come up with a pilot test case. The pilot will give a better picture of the expected output after scaling.
- HP Reliability
- James Kovacevic’s LinkedIn
- Reliability Report
- Eruditio Supports: www.help.eruditio.com
Fred Schenkelberg Links:
- Fred Schenkelberg s LinkedIn
- FMS Reliability
- Speaking of Reliability Podcast
- Accendo Reliability
- Soft Skills for Reliability Engineers
Rooted In Reliability podcast is a proud member of Reliability.fm network. We encourage you to please rate and review this podcast on iTunes and Stitcher. It ensures the podcast stays relevant and is easy to find by like-minded professionals. It is only with your ratings and reviews that the Rooted In Reliability podcast can continue to grow. Thank you for providing the small but critical support for the Rooted In Reliability podcast!