Maintenance practices are evolving in sync with the challenges facing the industry. For example, at first, the production needed effective means of dealing with accidents. So, reactive maintenance became preventive.
Later the focus shifted to asset health monitoring. This made it possible to reduce maintenance costs by reducing the number of checks. Helped automate the process and distribution of Industrial Internet of Things (IIoT) solutions. They also reduced the role of the human factor in the organization of maintenance.
Advanced data analysis has taken maintenance to the next stage of development – Predictive Maintenance (PdM). Predictive maintenance focuses on identifying patterns and predicting risks. This strategy has not only helped tackle disruptions that have not yet occurred but also changed the way we think about the efficient use of assets.
Today, the industry faces new challenges related to the scarcity of resources and energy carriers, the restructuring of familiar supply chains, and the introduction of new business models. This means that the principles of maintenance will change again.
Experts call Prescriptive Maintenance (RxM) the next step in improving maintenance management. This strategy is based on the development of practical recommendations (prescriptions) for the maintenance and repair service. By following them, you can reduce operational risks, increase the efficiency of actions or achieve the set KPIs.
The basis for Prescriptive Maintenance is Predictive Maintenance. While both methods focus on asset reliability and use similar data collection tools, there is a difference in approach.
Predictive maintenance predicts failures and helps you develop an action plan to prevent them.
Prescribing services allow you to analyze the effectiveness of the proposed plans in order to choose the best option among them.
In simple words: Predictive Maintenance answers the question “what can happen?”, And Prescriptive Maintenance – “what is better to do?”
It is important to consider the development of maintenance methods as an evolutionary process, and not to oppose RxM and PdM. The new strategy does not contradict the principles of predictive service, but only enhances its characteristic advantages:
– even less unplanned downtime;
– maximum performance of assets;
– profitability of service by increasing its efficiency;
– additional virtualization thanks to remote access to data;
– more integration with business process management tools.
But Prescriptive Maintenance also has a longer-term goal. In the future, IT systems will learn to not only create recommendations but also independently enforce maintenance orders.
Methods and tools
The main task of Prescriptive Maintenance is to provide the MRO with an optimal maintenance plan that will solve the maximum number of problems with the minimum number of actions.
So, for example, the manufacturer of elevator equipment TK Elevator, back in 2017, implemented a solution based on the principles of Prescriptive Maintenance. Their system analyzed the possible causes of failures and made recommendations in advance for the maintenance and repair service, which helped to eliminate 90% of potential problems during one team visit. Similar tools are also found in the arsenal of IBM, Airbus, and other service-oriented companies.
For the recommendations to be effective, accurate predictions are needed. And to improve the accuracy of predictions, advanced methods of multivariate analysis of anomalies are used.
It is necessary to do more than just pick out anomalies among the huge data stream that would indicate a potential problem. Anomaly should be detected as early as possible in order to create a maintenance prescription in time and minimize costs.
Cognitive IT systems are capable of this, they are able to analyze, interpret and adapt unpredictable information. For Prescriptive Maintenance, solutions are created that combine the advantages of Industry 4.0 technologies:
– Big Data;
– Artificial intelligence;
– Machine Learning.
The basis for the deployment of such solutions should be a powerful infrastructure based on IIoT. And while cognitive learning systems for RxM are still considered fantastic, tools for collecting and analyzing MRO data are already available. SmartEAM offers state-of-the-art products for digital transformation and the implementation of up-to-date maintenance principles in manufacturing.
You need it now
It might seem that Prescriptive Maintenance provides everything the same as Predictive Maintenance, but requires much more investment in the IT infrastructure.
This is not the case, as prescriptive care has a number of unique benefits that will soon become critical to the industry:
– using historical data on assets to calculate scenarios and create service templates;
– automatic launch of actions within the framework of the plan to ensure the reliability of the equipment;
– KPI forecasting, modeling and scenario assessment for a wide range of tasks;
– optimization of the service plan, taking into account real material and technical constraints.
These features allow enterprises to reduce the time it takes to find optimal solutions in the field of maintenance, gradually reducing the role of people and resources in optimizing MRO. And this is exactly the strategy that matches the future challenges of the industry.
Maintenance practices are evolving in sync with the challenges facing the industry. For example, at first, the production needed effective means of dealing with accidents. So, reactive maintenance became preventive. Later the focus shifted to asset health monitoring. This made…
The industry is gradually recovering from the effects of the pandemic, analyzing mistakes in risk management and looking for new mechanisms to counter the impending challenges. And there are enough of them: experts predict an increase in the deficit of…
According to McKinsey, equipment maintenance accounts for 10 to 25% of all operating costs. And these funds are not always spent rationally. Regardless of whether the enterprise implements maintenance and repair on its own or outsourcing, everyone encounters examples of…