Data Analytics for Reliability Engineers

Data Analytics becomes a key tool for the Reliability Engineer. Using Data Analytics, the Reliability Engineer can find patterns to improve. For me it is an increasingly important tool while facilitating complex problems

Want to know more about the impact of Data Analytics? Check This survey of the International Journal of Production Economics on the impact of digitalization in the maintenance department. Data Analytics is recognized with the most impact for the future. They argue that: "The value of data analytics will lie in the ability to identify patterns and root causes and take proactive action to avoid disturbances and failures, thereby increasing productivity".

What is my experience?

Most Reliability Engineers are currently working mainly in Excel and not yet with Data Analytics packages such as Splunk, R, Celonis or Disco. It's a matter of time until the successful Reliability Engineer also adds the skills of a Data Scientist to his toolbox.

The practice

Last week a customer showed me his application of Data Analytics. By analysing the parameters (temperatures, valve control, air flow, emissions) of the process, he was able to extract inconsistencies from the controller (PID) and thus optimize and improve the process. It doesn’t get more practical than this with Data Analytics.

To do this successfully 2 factors are crucial:

1. It was possible to analyse all data
2. He was able to interpret the data

Excel can handle a lot but not with a lot of data. The beforementioned packages are much more suitable for this. The moment you as a Reliability Engineer can work with these packages you can call yourself an Analytics Leader according to Gartner.

Machine Learning

Machine Learning allows you to make predictions with datasets so that you can increase reliability. Taking decisions is automated with Machine Learning (self-driving car) and has different levels (Wikipedia).

Back to Practice

The customer who showed me the Data Analytics took it a step further: he had created a Machine learning model with supervised learning that found small deviations that were reported before they were a problem for business operations.

The next step

Quietly waiting does not fit with CoThink and we go along with the innovations and developments. We have developed a training module about data analytics in Problem solving. With this training and our own experience, we are already helping several companies to predict problems based on big data.

Do you want to stay informed of our developments? Send me an invitation and I’ll keep you posted.