Preventive maintenance has long been the tool of choice for enabling the constant and efficient operation of plants and equipment. Today, however, more and more companies are relying on more advanced and efficient tools, which allow this concept to be surpassed in favor of predictive maintenance. This practice, in fact, grants greater flexibility, efficiency and, above all, the possibility of containing costs by intervening only when it is really necessary, predicting destructive events in a timely manner before they occur.
Preventive and predictive maintenance: the differences
Undoubtedly, preventive maintenance has long been a part of good business practices in all industries and particularly in manufacturing, where the efficiency of plant and machinery is a top priority. At its simplest, we can define it as a set of procedures and practices aimed at eliminating errors and malfunctions before they happen. In this way, the reliability of equipment and facilities can be improved. The SAP ecosystem allows it to be managed through planned maintenance that generates calls based on time windows or strategies.
However, preventive maintenance has two critical issues: since it is essentially based on “static” forecasts derived from general statistics, it cannot predict unusual or exceptional events. Moreover, precisely because it provides fixed time windows for maintenance work, these can sometimes occur too far in advance of the actual wear and tear or life cycle of equipment.
Predictive maintenance is its natural evolution. Using data collected from control instruments, particularly IIoT instruments, it is able to create a more accurate and detailed picture of likely events, their time windows, and the likelihood of their occurrence. To simplify, we can say that predictive maintenance allows action to be taken only when there is a real need to do so, but well in advance of errors, failures and breakdowns.
This is made possible by an advanced data collection, management, and analysis infrastructure, which enables the creation, for example through machine learning, of increasingly timely and reliable predictive models. SAP has specific solutions for predictive analytics, for example SAP Predictive Asset Insight.
How to start the transition from preventive maintenance to predictive maintenance
In order to start a transition process to predictive maintenance effectively, it is essential to take some key strategic steps, which we can summarize in three key points.
The first is the ranking of assets on the basis of both their strategic importance, economic value, and complexity. In fact, a phased approach tends to be to adopt predictive maintenance, which is more complex, on key assets, both in operational and economic terms. It will be possible to decide at a later stage whether to retain a preventive maintenance policy for less relevant assets or to implement predictive maintenance for these as well, even after a cost-benefit assessment.
The second step is to arrange for data collection and storage. Typically, the quantity and quality of data required for effective predictive maintenance requires an adjustment of the instruments on the machinery. Fortunately, today the cost of adding sensors and measuring instruments has been greatly reduced thanks to wireless and IIoT solutions. Identifying precisely what data is needed is often a task that requires specialized consulting efforts aimed at determining the most stable, cost-effective and effective solution.
The third step is to apply the analysis tools. Indeed, to be effective, predictive maintenance must be as asset-specific and optimized as possible.
For this to be possible, it is necessary to have data collection and analysis tools that are powerful, functional and, above all, configured in the right way. An additional aspect to consider is that data can be collected and used in a collective pool, homologous for all assets of the same type in order to consolidate the statistical base, but the analyses must necessarily be asset-specific. This introduces considerable levels of complexity, but these can be addressed today due to the increasingly low cost of computing power and storage space. Again, specialized consulting will be able to provide the necessary tools.
The benefits of predictive maintenance
If one were to engage in a process of oversimplification, one could observe how, compared to predictive maintenance, this new mode essentially meets the need that generated its genesis, namely to predict possible malfunctions before they happen. In this way it is possible to avoid interventions that, while essential in the overall economy of resource efficiency, are often redundant. Knowing the behavior of equipment in advance ultimately means precisely eliminating these redundancies in favor of a more efficient system.
Our RegestaLAB division offers asset management solutions for maintenance monitoring and prediction.
Contact us to receive information and to make your business a smart business.
