AI, machine learning, predictive maintenance

You Can't Fake Culture

October 07, 2020

There’s a lot of activity around digitalization initiatives, analytics and data science projects and dealing with the current work-from-home (WFH) situation and a remote workforce. As with any such project or program there is always the question: “How do I know this is working?”

Questions like this bring out the engineer in us and typically we develop some sort of key performance indicators (KPIs) to mark our progress. Now, I’m not saying these measures aren’t useful, but they do have a way of finding the sunny side of the story. After almost 40 years in and around process manufacturing, I have seen too many programs based on a “fake it ‘til you make it” strategy. I believe that the programs and initiatives that are really succeeding create changes in culture. And you can’t fake culture. 

When programs like this succeed, there are inferential indicators of success that are far more telling than any network of KPIs. So, I believe the key question to ask is “has this project, program or initiative impacted our culture?”

When culture changes, we see several things occur. The analytical results become ubiquitous in the reports, dashboards and performance measures for the business and its people. They become trusted sources of truth. When initiatives gain traction, we inevitably see new champions and evangelists rise in the organization. We also see new workflows that tear down conventional wisdom in favor of new methods. 


Examples of Digital Transformation Programs Changing Culture 

Recently, Saras, one of the leading independent players in the European energy and refining market, launched a highly successful digital transformation program. One component of the program is a Digital Predictive Maintenance Center. To ensure the program had the intended results, Saras focused on change management around both people and technology. 

Based on a successful evaluation where staff saw how quickly a new predictive maintenance solution delivered value, the company deployed the software to monitor key refinery assets. As teams saw how the new solution accurately predicted failures in advance, they began to trust the tool. Saras then trained staff to manage the software, allowing the company to independently maintain, deploy and scale the solution. Saras began using the predictive maintenance solution on assets in a wind farm, as well as in the refinery, avoiding between €4M and €5M in maintenance costs and lost production in less than two years. 

Digital transformation isn’t really new; decades after Deming took the control chart to post-war Japan, online statistical process control (SPC) became ubiquitous in manufacturing as a way for operators to track the process and find problems. Another example of successful early digitalization is advanced process control (APC) where dynamic matrix control technology completely transformed process management. Both examples gave rise to entire communities, mavens and lexicons.

KPIs and the drama they catalyze are an unavoidable fact of life. But with any measurement comes the need to understand the accuracy and repeatability of the instrument. Certainly, do not ignore the directional information from these calculations but backstop them with supporting evidence that can’t be faked. 


For additional information, read the case study, Saras Drives Innovation in Predictive Maintenance via Digital Transformation Program.



There was a problem storing your subscription

Leave A Comment