Elipsa
Published in

Elipsa

The Two Sides of Predictive Maintenance

Intelligent Monitoring Pre and Post-Production

Predictive Maintenance

Before diving into what we mean by the two sides of predictive maintenance, let’s start by saying what we mean by predictive maintenance in general. A lot has been written about predictive maintenance, some enlightening and some simply just filled with buzz words around industry 4.0.

When we discuss predictive maintenance, we are referring to the ability to use machine learning to find patterns in historical machine data in order to teach the system what normal operations look like. Quite simply, we look to “teach” the system what normal looks like so that it can monitor the same machine in real-time and detect when something looks abnormal. The idea behind this is that this abnormality is indicative of a future failure. With this alert in hand, companies can plan downtime, perform suggested preventative maintenance, order necessary parts, etc.

The goal of predictive maintenance is to use AI to better understand and maintain equipment in order to increase its useful life.

At Elipsa, we have built a no-code solution to build and deploy such predictive maintenance models without the need for a data scientist.

The Two Sides

So, when we say there are two sides to predictive maintenance, what we mean is that there are two distinct places that the same technology can be used in the product lifecycle. Each utilizes the same machine learning techniques but offers its own unique value proposition.

Production Phase

Let’s say you work in a factory that produces widgets. These widgets pass through a series of machines on the factory floor as it moves from raw material to work in process to finished good. These stages of production each perform an integral part in the manufacturing process and each presents a potential point of failure and risk.

If a single machine in the process goes down, it could mean halting the entire production. This downtime can lead to a significant amount of lost revenue as well as significant repair costs.

(check out our post on how AI positively impacts the cost of quality)

Photo by Science in HD on Unsplash

As IoT expands in the industrial world, machines are being fitted with new sensors measuring everything from vibration to the RPMs of the motor. Artificial Intelligence, in the form of machine learning, can take the collection of data from a machine and find patterns beyond what an engineer or any human is capable of. These settings can be monitored in real-time to get a better understanding of the machine’s health.

Beyond that, every machine is different. As a result, you often need multiple AI models in order to monitor your entire facility. Through elipsa’s “clicks not code” approach, companies can easily build and deploy a series of AI models that connect back to your existing systems in order to monitor your entire operations with ease.

Predictive maintenance can lead to higher Overall Equipment Effectiveness (OEE) by ensuring less downtime and greater productivity in the production phase.

Post-Production Phase

Oftentimes, many who talk about predictive maintenance stop at the manufacturing phase. However, as more and more machines on the factory floors are capable of adding sensors, so too are the machines that you and your company are producing.

The widgets that you produce in the factory may be a single part that is used as material in another manufacturing process. But, if your widget is the end product such as industrial machinery, medical equipment, etc then there is a growing chance that your product includes sensors as well.

Photo by Dominik Vanyi on Unsplash

According to the American Society of Quality, 60% of the cost of quality is from defective or poor quality products that reach the customer. In other words, costs due to defects, recalls, and repairs are the highest quality-related cost to the bottom line. This is particularly true when it comes to warranty claims.

Deploying predictive maintenance on products under warranty in the field can allow a company to intelligently monitor the operation of each and every machine. Similar to the machines on the factory floor, this new form of monitoring can allow a manufacturer to better understand the real-time state of every product they have under warranty allowing them to get ahead of catastrophic problems by ensuring the health of their machines.

Deployment of predictive maintenance could allow a company to lower their warranty claims and long terms liabilities but it could also lead to increased revenue. Manufacturers could utilize predictive maintenance as part of new IoT applications built around selling extended warranties to their customers helping to further extend the machine’s useful life while generating recurring revenue in the process.

Conclusion

As we say, predictive maintenance has two sides but really what we mean is that it can be deployed at multiple stages of a product’s lifecycle. Each enhances the monitoring capabilities of the company, helping to decrease downtime inevitably decreasing cost, while increasing quality, revenue, and customer satisfaction.

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store