“If it ain't broke, don't fix it!” the old saying goes, but the rise of predictive analytics is turning that idea on its head by helping manufacturers correct issues before they arise. Welcome to the age of "fix it before it breaks."
Today, poor maintenance strategies can reduce a manufacturers overall productive capacity by 5 to 20%. Recent studies also show that unplanned downtime is costing industrial manufacturers an estimated $50 billion each year. This begs the question, how often should a machine be taken offline to be serviced? Traditionally, this dilemma forced most manufacturers into a trade-off situation where they had to choose between maximising the useful life of a part at the risk of machine downtime (run-to-failure) or attempting to maximise uptime through early replacement of potentially good parts (time-based preventive maintenance), which has been demonstrated to be ineffective for most equipment components.
But thankfully, today’s manufacturers don’t need to compromise thanks to the introduction of smart manufacturing. The industry is adding new capabilities to operational technology, including remote management and operational analytics to gain better insights in their production lines. The number-one value-add has been predictive maintenance.
Predictive maintenance involves collecting and evaluating data from machines to increase efficiency and optimise maintenance processes. This meaning you can gage the condition of equipment and also more accurately predict when maintenance work is needed before issues occur. Not only does this mean scheduled downtime could be a thing of the past, but there will also be no need for manual inspections, saving both valuable time and money for today’s manufacturers.
For many manufacturers, this technology may still be on the wish list, but let’s take a look at how predictive maintenance is working in practice across a range of common machines in factories around the world:
Robots are an integral part of most manufacturing plants these days but they do need frequent maintenance checks. It’s difficult to plan robot maintenance if the health of a robot is monitored only locally or not at all. But why refrain from gathering relevant machine data? Many parameters can be monitored, including CPU and housing temperature as well as positioning and overload errors. By collecting and displaying this data centrally and then evaluating it, maintenance can be planned before issues arise.
2. Heat Exchangers
Pharmaceutical goods, packaging, food processing and oil and gas manufacturers may all use heat exchangers.
This crucial piece of machinery can easily become clogged due to deposits in the conduits. A further complicating factor is the fact that it is impossible to measure the flow rate of a heat exchanger directly. A complete blockage can cause serious problems, resulting in manufacturing errors and hours of downtime.
One solution to this issue is to measure the temperature differential upstream and downstream of the heat exchanger. After gathering and visualising the measured values, it is possible to define threshold values. These values can then be input into an alert system to notify employees as soon as the first signs of clogging appear.
3. Milling Machines
Milling machines are used to drill, bore, cut gears, and produce slots. Spindles in milling machines are prone to breaking during the production process and repairing spindles can be very expensive. Therefore, being able to predict damage and precisely when the spindle will break can greatly reduce costs.
To overcome this challenge, special sensors such as ultrasonic or vibration sensors identify the patterns of a fragile spindle. Relevant alert settings for the current state of the machine can then be created to alert the operator of when the spindle is likely to break.
Making Predictive Maintained a reality
Although some manufacturers are already beginning to adopt IoT solutions, we are still in the early stages of the ‘smart factory’. Most organisations are still focused on collecting a wide range of big data and establishing the correlations between different ways of working and product quality. Once they have enough data, businesses will begin to create increasingly sophisticated models that more accurately predict failure.
To unlock the full potential of predictive maintenance, organisations must become more adept at managing the growing volume of data within their organisation and ensuring it is fit for analysis. An analysis might include data from spreadsheets, databases, social media and even photos, so having the right data preparation processes in places will be crucial to combing all this information in a cohesive way.
ANS has been working with manufacturers to bring predictive maintenance capabilities to reality. With end to end capabilities from sensor to action. ANS can deliver IoT solutions delivering comprehensive dashboards, showing the location of items, movement information and current state of assets tracked through the IoT solution. Managed IoT solutions that provide real time data analytics, event based warnings and machine learning capabilities to predict outcomes.
ANS have partnered with the world’s largest suppliers and manufactures of sensors, able to utilise long-range, low-power WAN technology, or mobile data connectivity, coupled with enterprise grade cloud computing platforms and managed application development ANS enable you to derive operational insights from your people, assets and customers with ease.
While we believe the potential of big data in maintenance has yet to be fully realised, there is one prediction we can make for sure: predictive analytics is going to become an increasingly important part of how manufacturing plants are run and ANS will be at the core of delivering these services to help you innovate and operate better, faster.
Posted by Nathan Hopkins