Sunday, September 4, 2022

Game Changers - Predictive Maintenance

 Introduction

Human beings have come a long way in a relatively short period of time, it was not so long ago that we were actually keeping track of maintenance actions on a log sheet.  Inventors see the world in a different way, not just as it is, but what it could be.  Often times innovative ideas come from not just thinking outside of the box, but also what we can do with the box.  The sky is not the limit, instead, the sky is our playground.  Throughout history, we humans have been exceeding the status quo and pushing the limits, from the wheel and plow, to vaccines and medicine, satellites, smartphones, and even unmanned airplanes.  So, what is the next step how will science and technology continue to evolve and revolutionize the world as we know it today?  To answer this question, we must consider how inventors become innovative, what motivates them, and where do they get their ideas from.  Gaming ideas usually come from errors or unwanted accidents that have occurred, including predictive maintenance.

History of Predictive Maintenance

    The history of predictive maintenance began in the early 1950s, and it was primarily developed by the Japanese (Roser, 2021).  Back then maintaining your equipment and machinery was essential to your business.  Often times maintenance was more reactive than it was proactive, the motto was simply if something breaks then fix it.

    After World War II the Japanese economy and industrial infrastructure was in bad shape.  At that time the Japanese leaders began to adopt American philosophies about quality management and control.  This led to Total Quality Management and was introduced by Dr. W Edwards Deming during a lecture in Japan.  More lean principles were introduced and there was more of a widespread emphasis placed on automated equipment, tools, processes, and procedures.  As we moved to a more digital platform that greatly increased the efficiency of production, this also brought along many unforeseen challenges.  As skilled workers diminished, there were not enough technicians that could maintain the machines, and this led to costly breakdowns.  This contradicted their new paradigm to produce high-quality products, you cannot achieve high-quality products with poorly maintained equipment (Cumming, 1950).  This spearheaded predictive preventative maintenance, which is the concept of daily maintenance.  This consist of properly managing and planning maintenance activities to keep the equipment in good operational condition.  The intent is to avoid failures and unwanted accidents by engaging in periodic inspections, routine maintenance, and prevention of deterioration. 

    Toyota was one of the first companies to introduce preventative maintenance, by introducing automation, then by increasing maintenance technicians, and also by implementing preventative maintenance techniques (Roser, 2021).  This helped to advance technology as we know it today, these combined techniques and practices with earlier lean principles is what helped to lead to unprecedented levels of efficiency and success.

Predictive Maintenance Today

The intent of predictive maintenance is to improve maintenance efficiency and effectiveness, see figure 1.  Technological advancements were made to ensure that the maintenance actions were carried out, but also performed in such a way that it is cost-effective.  It was important to factor in proper maintenance planning and scheduling, while also ensuring that lean maintenance principles were followed (Wireman, 2004).  Unwanted accidents and catastrophic events led to predictive maintenance solutions and failure reporting systems that are supported by machine learning. 

Figure 1. Predictive Maintenance and Its Role in Improving Efficiency

Companies have now turned to predictive maintenance to avoid these severe economic losses and to increase system reliability.  With the use of machine learning, companies can implement predictive maintenance techniques.  Data has been developed and collected, and preventative maintenance solutions have been tested in a simulated environment, and the results demonstrate that preventative maintenance technology is able to predict machine states and failures with high accuracy (Paolanti et al, 2018).

There are also economic forces that are driving predictive maintenance because it helps to drive cost efficiency.  Predictive maintenance can drive substantial savings by increasing production, performance, and quality.  This will also lead to cost savings with reduced maintenance actions because innovative technology now allows us to predict failures before they happen.   

Before predictive maintenance existed, there were several accidents that transpired.  Machinery and complex systems like aircraft were inspected based on handwritten planned-out maintenance schedules, which were determined by the manufacturer based on the life expectancy of parts studies.  In this case, some parts of the aircraft were required to be disassembled so that those sub-parts could also be inspected and analyzed to ensure proper maintenance.  There was always a risk associated with the inspection activity because it could actually introduce failures if the inspector failed to put the sub-parts back together correctly.  So this required a new approach that required that maintenance actions actually be in sync with the current condition of the equipment and that it should be carried out based on the equipment’s actual usage patterns.  Based on the usage patterns, data could be collected, and machine learning could be implemented to predict preventative maintenance actions that should be taken in order to prevent unintended accidents.  This was the true beginning of innovation, which led to the use of probability statistics to determine inspection cycle strategies which is ultimately the foundation for predictive maintenance, which is a driving technological force.  

Figure 2. Predictive Maintenance Infrastructure


References

Cumming, W. (1950).  The Origin of Preventive Maintenance And What It Includes,"

    SAE Technical Paper.  https://doi.org/10.4271/500032

Paolanti, M., Romeo, L., Felicetti, A., Mancini, A., Frontoni, E., & Loncarski, J. (2018, July).

      Machine learning approach for predictive maintenance in industry 4.0. In 2018 14th IEEE/ASME         International Conference on Mechatronic and Embedded Systems and Applications (MESA) (pp. 1-        6). IEEE.

Roser, C. (2021).  A Brief History of Maintenance.  Retrieved on September 4, 2022

    from https://www.allaboutlean.com/maintenance-history/

Wireman, T. (2004). Total productive maintenance. Industrial Press Inc. Global

    Education US. https://coloradotech.vitalsource.com/books/9781119713197

No comments:

Post a Comment

Conceptual Framework