The Importance of Predictive Maintenance
Published on : Monday 30-11--0001
“In the long run, all machines break down” is admittedly a terrible reinterpretation of a famous quote by John Maynard Keynes, the renowned British Economist. However, there is a kernel of truth in it. All machines are prone to wear and tear. In fact, a joint study by Wall Street Journal and Emerson found that unplanned downtime costs industrial manufacturers an estimated US $50 billion per year. Equipment failure is the cause of 42% of this unplanned downtime.
So, what can we do to prevent this colossal loss?
The answer is simple – Employ a more efficient asset maintenance strategy.
In the early days of the industrial revolution, when machines were first introduced it had a huge impact on productivity. Production not only became faster but it also became safer. Later with the assembly line, the entire production was broken down into simple manageable tasks where workers would use standard components to complete a portion of the product before moving it down the line. The assembly line led to mass production as we know it today. In fact, by using the assembly line, the Ford Motor Company was able to produce a Model T in 90 minutes flat.
This march towards higher productivity and efficiency continued into the age of electronics with the introduction of the Programmable Logic Controller (PLC). These PLCs were instrumental in getting rid of racks and racks of relays that were originally used to control machines. The PLCs brought automation to the factory floor and ushered in the third wave of the industrial revolution. Today, we are at the onset of the 4th industrial revolution also known as the Industry 4.0 or Industrial IoT. Regardless of the terminology that one wishes to use, this trend is all about using sensors, intelligent devices, inter-connected systems and analytics to improve decision-making which in turn will lead of better production output and lower costs. It is estimated that Industrial IoT will allow manufacturers to increase their productivity by 30%.
Industrial IoT is also helping manufacturers transform their businesses as well as enhancing customer experience through new engagement models. For example, the Airline industry is now driven by the “Power by the Hour” model where Airlines pay a fixed sum per flying hour for a jet engine instead of buying it outright. These kinds of “product-as-a-service” models are becoming more and more popular. In addition to the new revenue streams that manufacturers are able to generate, Industrial IoT also helps to improve operational efficiency through enhanced worker safety, improved productivity and better asset management.
In fact, optimised asset utilisation is increasingly seen as a competitive advantage in today’s environment. Factories have started measuring and closely tracking various performance metrics related to their assets including production output, overall equipment effectiveness, personnel productivity etc. Maintenance of assets, which was seen as an activity to be undertaken only when there was a breakdown – also known as Reactive Maintenance strategy -has become much more important. Today, most maintenance strategies range from Periodic to Proactive to Predictive.
Figure 1: Maintenance Strategies.
The maintenance strategy that employs advanced analytics to predict machine failures is known as Predictive Maintenance. In a survey carried out by the World Economic Forum (WEF), the most widely cited application of the Industrial IoT is predictive maintenance and rightly so. Predictive maintenance allows manufacturers to lower maintenance costs, extend equipment life, reduce downtime and improve production quality by addressing problems before they cause equipment failures.
However, to realise the benefits of predictive maintenance, one must overcome a few challenges around the lack of data and lack of digitization. Factories have different levels of maturity when it comes to data collection and analysis. A lot of information is still paper-based and that obviously needs to change if we want to use advanced analytics. A robust digital infrastructure must be put in place which includes connectivity with assets, the deployment of an IoT platform like the Microsoft Azure IoT Suite or PTC Thingworx and the use of statistical techniques such as machine learning to analyse machine data. Predictive analytics helps us to predict future outcomes using past data. But, it is important to understand that predictive analytics is a journey and not a destination.
It starts with identifying the right asset to manage. For some assets, periodic maintenance is the right strategy. However, for critical assets, it is important to put together a predictive maintenance strategy. The methodology for identifying critical assets is called Equipment Criticality Analysis (ECA). The ISO 55000 standard for Asset Management defines a critical asset as an “asset having a potential to significantly impact on the achievement of the organisation’s objectives”.
Once the critical assets are identified, they need to be connected to the digital infrastructure so that real-time data can be ingested for remote monitoring. Data preparation and data quality are the key inputs for any predictive model. The more high-quality data we are able to feed into the predictive model the better its accuracy. The efficacy of the model is first tested offline before it is integrated into live operations and it keeps improving over time as more failures are accurately identified and tagged. All this may seem a little daunting since analytics is not a core function within a manufacturing setup, but there is help out there from companies that provide IoT technology solutions and engineering services.
As a manufacturing unit, one can initiate the predictive maintenance of assets through a Proof of Concept (PoC) in partnership with a technology provider. The scope of the PoC may be small, but it is important to evaluate the scalability, security, enterprise integration capabilities of the solution. Also, while it is good to look at the technology aspects, one must not ignore the human element in all this. It is very important to empower employees and workers through workshops and trainings. After all, it is the employees who will be the users of any new technology solution. A cross-functional team of employees should be involved in putting together new strategies and processes. Finally, continuous improvement should become ingrained into the culture of the organisation.
According to a study by McKinsey, predictive maintenance can help reduce maintenance costs by 10-40% as well as reduce equipment downtime by 50%. This will allow manufacturers to produce high-quality products faster at lower costs thereby benefitting both the producer as well as the consumer.