Whether it’s about temperature, pressure, vibrations, oil levels, or wear—these days organizations can measure so much with sensors in their installations. Based on the measurements, for instance, technicians can predict what type of maintenance is needed, and when. Because sensors make it possible to predict the right maintenance moment, reactive maintenance becomes largely superfluous. It is certainly not the case that every organization will be using this data analysis everywhere within the next couple of years, but it will quite probably be universal within the next decade.
From reactive to proactive maintenance
Today, many organizations still perform reactive maintenance on their capital equipment. This type of maintenance only occurs when there is a failure. It is always unplanned, unexpected, and mainly, expensive. After all, when a device or operational piece of equipment really is defective and has to be repaired or replaced, it can be assumed that important business processes will be halted for longer than desired. Consider, for instance, a flickering fluorescent tube in a meeting room; perhaps employees are unable to use this area for some time. To avoid reactive maintenance (and high costs), maintenance departments create preventive maintenance plans, based mostly on time. That means the fluorescent tube in the example may be replaced earlier than needed.
In manufacturing, there is a heavy emphasis on proactive maintenance. Here, sensor data helps to predict the right maintenance moment, thus minimizing undesired disruptions. This development is carrying over into the entire maintenance domain. Now that the use of sensors is becoming steadily cheaper and more accessible, sensors are being built in by default into (new) installations, and the technology for data analysis is available, it becomes possible to switch to “just in time” proactive maintenance. This results in more control over the performance of the organization’s operational equipment and a more efficient deployment of available maintenance resources. With this strategy, maintenance can be performed at times when it does not directly cause serious problems in the business processes.
Will maintenance plans continue to exist?
There was a time when an experienced technician could determine if maintenance was needed just by hearing a particular noise; however, with today’s complex machines, that is an impossible task. The power of sensors is that data is collected continuously. In combination with big data analysis, patterns can be recognized so that fault and disruption predictions can be made. That enables organizations to take action in good time because there’s an insight into how an operational resource is used, how it is stressed, and what the environmental factors are. In the years to come, this scenario for maintenance management will become the rule rather than the exception. Drawing up maintenance plans in advance, a time-consuming task, will then become less necessary.
In the classic way of performing maintenance, the ratio of 70% planned and 30% reactive maintenance will only be achieved by “best in class” organizations who have their maintenance process properly ordered. In all probability, the deployment of sensors will cause a shift, whereby reactive maintenance will become obsolete. The best possible maintenance moment can after all be predicted, because it is calculated on the basis of historic data and the current status of operational resources. Perhaps the use of sensors sounds like sci-fi right now, but before you know it, that’s what you’ll also be working with!