Why AI Fails with Matt Kirchner

Matt is the Chief Product Officer at Atonix Digital and is involved with AI, works on software packages, analytics etc. More about Matt:

He has a background in mechanical engineering and started off in a large firm where they offered remote monitoring services to power generating customers. They oversaw reliability and operational performance. Over time, the company started investing in software tools to facilitate better and more efficient monitoring. The firm then split and now a portion of it only focuses on operational intelligence tools.

Despite such progress some companies still lack basic data collection tools;

Certain industries are further along the path of adoption of these technologies but even within these industries, some companies are ahead of others.

Of importance is effectiveness in continuously improving.

 

What is operational intelligence and how is it different from reliability or preventive maintenance programs?

Reliability programs can be perceived in three ways;

1. Programs focusing on the details of what is happening in an equipment. Such data is high frequency and focuses on detailed analytics.
2. Programs based on medium frequency data which are the actions and activities in a process. I.e. Process data
3. Longer term maintenance planning based on such factors as the age of the equipment.

The difference is based on the data sources.

Operational intelligence obtains data from activities and actions in the process line (medium frequency data) used to make near-term decisions.

 

Why do many organizations miss operational intelligence in reliability?

Because of the overselling of advanced analytics, machine learning and AI such that there is too much expectations and overreliance on these programs.

Organizations also try to deploy all these programs at once without a proper work management program.

 

How do companies benefit from operational intelligence?

• It helps in planning equipment preventive maintenance.
• Aids in monitoring process performance by analyzing operational costs such as water and energy consumption.
• Gives a forewarning before failures become immediate safety concerns or even emergencies

 

Why are so many organizations unable to implement operational intelligence?

There are several reasons including;

• The way AI and other programs are marketed bombards potential users with many complex alternatives that it creates confusion.

Analytics should target the experts and should be simplified for easy consumption and implementation.

• Many users expect the program to do everything. This is wrong because the calculations will only point a user in the right direction but they must choose how to act.
• People tend to overthink the programs and expect to achieve immediate perfection instead of taking small incremental steps towards success,
• Implementers might have too narrow or too broad focus in the pilot phase; if the focus is too narrow, perhaps on one equipment, then results may not be achieved. It could have been achieved if a few more pieces of equipment were within the scope.

Adopting a too broad approach at the same time at the pilot phase creates difficulty in monitoring a new program. Pick a site or a major process for the scope of a pilot phase. However, do not perceive the pilot phase as experimental or a step for proof of concept. Instead, explore a hypothesis for a program that you already know will work.

 

How do organizations start implementing operational intelligence?

• Pick an area to focus on for the pilot phase
• It is better if you already have data on the performance of the program.
• Use designed templates developed from prior knowledge even from service providers.
• Find a partner who is experienced in implementing operational intelligence.

 

What are the common characteristics of organizations that are successful in implementing operational intelligence?

Organizational culture is imperative i.e. there are those that are innovative and seek to do it themselves while others want to learn from those who have implemented these programs.

Operational intelligence can succeed in both environments such that; those who are not innovative can hire experts and receive training on how to optimize the program. Those who seek to do it themselves would still benefit from a 3^rd party advising the team on how not to disrupt their process during execution.

 

Why is it that some organizations have invested heavily but have not seen ROI?

Reasons include:

• Unrealistic expectations on analysis without actions.
• Mismatch of the tools and the resources
• In some cases, the analytics is superior and unusable; you need an environment where the analytics is powerful but usable in the process.

 

Last thoughts for anyone seeking to implement operational intelligence.

• Ignore the noise that makes the program appear excessively complex.
• Have realistic expectations on the program
• Find someone who is more experienced in the program to assist.

 

Key takeaways

• Do not be overwhelmed because the program is not as complex as misconstrued.
• Analysis without actions is useless.

 

How can people find you?

Atonix.com

 

Best resources:

Analysis reports from product or vendor reviews give good insights on the programs.

 

Eruditio Links:

• Eruditio
• HP Reliability
• James Kovacevic’s LinkedIn
• Reliability Report

Matt Kirchner Links:

• Atonix LinkedIn
• Matt Kirchner LinkedIn
• Atonix.com

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!