As machines and technology evolve, maintenance strategies evolve too. The newest maintenance strategy is prescriptive maintenance (RxM). Previously, the most advanced maintenance strategies were condition-based maintenance (CBM) or predictive maintenance. Prescriptive maintenance starts like predictive maintenance, but it uses machine learning and artificial intelligence (AI) to go above and beyond.
What is Prescriptive Maintenance?
There are three key stages that make up prescriptive maintenance:
- Use of sensors or IoT (Internet of Things) connected devices to consistently collect data on asset health and performance.
- Sensor and IoT data are analyzed to offer predictions based on current asset conditions. Predictive maintenance would stop here with an alert to the maintenance team to inspect or repair the asset.
- Prescriptive maintenance takes this data and uses a custom-built algorithm to detect the potential outcomes of the issue or failure. Then, it creates possible solutions and can even be automated to activate the best solution.
We’ll explain more with an example. You have sensors on a piece of equipment collecting various data points, including temperature readings. Temperatures begin increasing, and equipment failure is inevitable. Predictive maintenance analyzes the data and enables you to predict when that equipment failure will be. Prescriptive maintenance takes that prediction a step further by analyzing various solutions and outcomes. It tells you that the best solution is to slow down the equipment by X%, which allows you to get double the remaining useful life from that component. If you allow your prescriptive maintenance program to automatically act on solutions, it can make the speed adjustment for you and send the alert for later repair.
Is Prescriptive Maintenance Right for You?
Every maintenance strategy has its pros and cons. While prescriptive maintenance is the latest and greatest, it may not be a fit for every business and every asset.
Pros:
- You benefit from the highest level of automation using machine learning and artificial intelligence.
- The algorithm saves you time by providing multiple outcomes, which can even include forecasting the financial impact of each solution.
- Advanced sensors combined with strategic data analysis give you more support for challenging equipment failures or for undetectable equipment problems.
- Digital models of your current equipment enable you to create digital renderings of additional assets or equipment changes before you make a purchase or spend time making changes.
- As with any maintenance strategy, it reduces downtime, increases efficiency, and further improves your maintenance plan.
Cons:
- You may have some sensors already in place. But, to fully implement prescriptive maintenance there is an upfront investment in the necessary artificial intelligence, machine learning, and sensors. The cost is the number one deterrent for many companies.
- The algorithms are custom to your business and how you use the assets. So, the machine learning platform has to collect enough data to create the algorithms needed to be effective. This takes time and delays the return on investment.
- With any new technology, there can be obstacles to implementation. You may face resistance or hesitation around allowing AI to have unsupervised control of assets or systems that are mission-critical or can cause a safety incident. As the technology continues to improve, it will be more widely trusted and accepted.
Even though one of the cons is that prescriptive maintenance is a new technology, it is already in use by many industries from oil and gas to aerospace and aviation. For example, ThyssenKrupp uses prescriptive maintenance for their elevators. Their machine learning system has been so successful that it can predict equipment failure due to door problems five days in advance, automatically schedule a technician, and provide the top four repair solutions. This prevents breakdowns while people are in the elevator, and technicians fix the issue with 90% accuracy.
The Future of Maintenance
If prescriptive maintenance isn’t right for you now, it may be later. The technology will evolve and improve, which will eventually lead to a lower entry cost. Some companies see the benefit now and are choosing to implement prescriptive maintenance for their most expensive to repair or replace assets. However, if you’re just starting to explore digitizing your maintenance, then prescriptive maintenance is a big step. You can start with something smaller, like adding sensors and implementing Enterprise Asset Management (EAM) software. Then, you’ll be ready when prescriptive maintenance is more cost-effective and readily available.
