Predictive maintenance: from condition monitoring to prognostics
Condition monitoring in place - check.
Seeing the benefits - check.
Ready to take the next step - check.
But what is the next step in the world of predictive maintenance? The answer: prognostics. Prognostics is the science of forecasting when your assets will stop being able to perform their intended functions. With prognostics in place you can properly perform predictive maintenance – and it is undoubtedly the future of condition monitoring.
Condition monitoring (CM) provides a picture of machine health. It monitors a machine’s condition by checking sensor readings against predefined thresholds and raises an alert when these readings (such as vibration or temperature) fall outside of normal operating parameters. These alerts identify problems early, flagging any issues before the machine fails. CM allows maintenance to be planned and scheduled rather than being carried out sporadically and reactively, reducing unplanned downtime and calendar-based service costs, and extending asset life of machines through the use of ongoing personalized maintenance addressing issues before they reduce the potential lifespan.
Prognostics goes beyond CM by forecasting when an asset will stop being able to perform its intended function, giving you its Remaining Useful Life (RUL). Prognostics can be used on any machine, although it’s best when looking at collections of assets that have strong commonality. It is achieved by understanding the signs and characteristics of past failures and mapping those to current machine behaviors, recorded by various sensor inputs (ideally using data that you already collect for historical analysis).
The benefits of prognostics are huge, the biggest being that it can prevent machine downtime and increase machine lifespan due to a better understanding of machine health, rather than using a benchmark lifetime calculation.
The ability to capture and analyse all of this information is made possible through the increasing adoption of Industry 4.0. Predictive maintenance tools can be used by very good maintenance engineers – but to use it at scale, it’s a task best left to large computers able to run a number of algorithms simultaneously and use machine learning to understand performance.
The future: beyond CM and prognostics
In addition to their intrinsic potential, CM and prognostics are also enablers for further business transformation. The Rolls Royce ‘Power by the hour’ concept revolutionized the aerospace industry by shifting the offering from engine sales to delivering flying hours. Predictive maintenance is an enabler for this ‘servitization’ model, which could see manufacturers move from selling products to delivering products as a service. Could the term PaaS become as commonplace as SaaS? Time will tell.
Condition monitoring and prognostics look to make factory maintenance and management more efficient and more cost effective – and they have the potential to transform business models from a product to service offering. While CM is recognized as an effective maintenance and business solution, continuous development towards a predictive future holds huge potential.