The History of Predictive Maintenance

“You can’t fix something that isn’t broken” used to be the philosophy of traditional Reactive Maintenance, many years ago. Help was only called for once a machine had completely failed.

Completing reports, locating or ordering replacement parts and assembling the required maintenance expertise, often introduced unnecessary delays and downtime, as well as overtime costs. 

Lead times and, consequently, delivery schedules went out of the window. 

Planning around failure

Thinking was updated in the 1980s, with the popularity of Japanese manufacturing techniques including Just in Time (JIT), where the new idea was to conduct Scheduled, or Planned, Preventive Maintenance. This concept was further refined by the introduction of Total Productive Maintenance, used in Lean Manufacturing, where maintenance was planned to keep equipment in working order at all times. 

Successful implementation meant fewer breakdowns and reduced downtime, resulting in more effective use of labour, longer machine life and improved workplace safety. 

These obvious benefits improved overall productivity – but planning the maintenance involved assimilating information from equipment manufacturers, experienced engineers and operators, over a long period of time. 

->Download our free resource to learn the role of your IT team to be able to  deliver value and the best practices for data collection and transfer.

How could anyone be certain that maintenance was scheduled at the optimal time? 

Predicting Failure

Optimisation of maintenance timing has been achieved through Predictive Maintenance, by the introduction of Condition Monitoring. 

With this methodology, data from certain critical indicators is used to demonstrate when the equipment is showing decreased performance, or is due to fail. It serves as an early warning system, allowing time to plan for the contingency, reducing losses due to maintenance delays. 

The success of Condition Monitoring depends on obtaining quality, relevant information from these critical components in good time, analysing and evaluating it quickly, before acting swiftly to implement the findings. 

In traditional factories, limited information is available from industrial control systems, often requiring further processing and formulation into reports before experts are able to interpret the results. 

Now at your Fingertips

With the arrival of Industry 4.0 and the Smart Factory, every possible piece of data required to automate the analysis of the equipment’s condition can be in place. Digitalisation means that real-time data can be collected from multiple sources across the factory floor and stored in an easily accessible manner. 

This allows advanced analytics to be applied, quickly and easily, providing near-instant assessment of the situation and highlighting any anomalies. Processing in the cloud makes this information available across entire organisations, allowing them to pool best-practises and respond rapidly. 

The ability to auto-diagnose failure is further extended by the addition of Prognostics. Data gathered from the machines and assets build a model of the operation that, when analysed, provide advance indication of when a machine will fail, identifying its Remaining Useful Life (RUL). 

This anticipation of maintenance requirements allows for a more agile operation, with a well-understood time-horizon, allowing decisions to be made and actions to be taken that will prevent or limit loss due to unplanned downtime. 

Tell me more

Senseye automates the latest condition monitoring and prognostics techniques and displays things in a really simple way to help you to avoid unplanned downtime.

Download Effective condition monitoring: An enabler for Predictive Maintenance white paper