Data preprocessing is necessary to clean the data and convert it into a form from which you can extract condition indicators. If you don’t have quality data feeding into your machine learning model, the resulting predictions will be useless. You must store the data, clean it, integrate it with other data, and then analyze it for meaningful insights. #Udacity combining predictive techniques task3 seriesHaving enough data is great, but it’s just the first step in a series of steps for predictive maintenance algorithm development. Classification models to predict failure within a given time window.Regression models to predict remaining useful lifetime ( RUL).There are multiple modeling strategies for predictive maintenance and we will describe two of them (that I worked on the most) concerning the question they aim to answer and which kind of data they require: Once we have all this information, it becomes possible to decide which modeling strategy fits best with the available data and the desired output. The questions above should be answered by both domain specialists and data scientists. Since the operational life span of production machines is usually several years, historical data should reach back far enough to properly reflect the machines’ deterioration processes.Īdditionally, other static information about the machine/system is also useful such as data about a machine’s features, its mechanical properties, typical usage behavior, and environmental operating conditions. We also want information about maintenance and service history. Usage history data is an important indicator of equipment condition. Therefore, it makes sense to start by collecting historical data about the machines’ performance and maintenance records to form predictions about future failures. With Predictive Maintenance, for example, we’re focused on failure events. Predictive maintenance will detect the anomalies and failure patterns and provide early warnings.īased on my experience, the success of predictive maintenance models depend on three main components: Predictive maintenance avoids maximizes the use of its resources. Hence, it is not an optimal solution from a cost perspective. Although regular maintenance is better than failures, we often end up doing the maintenance before it’s needed. The cost of failure is much higher than its apparent cost.Ĭurrently, most companies deal with this problem by being pessimistic and through precise maintenance programs to replace fallible components before failures. If the engine of the bus breaks down, the company needs to deal with unhappy customers and send a replacement. To understand the importance of maintenance, let’s consider a bus company. In industrial AI, the process known as “training”, enables the ML algorithms to detect anomalies and test correlations while searching for patterns across the various data feeds. In contrast, ML algorithms are fed OT data (from the production floor: sensors, PLCs, historians, SCADA), IT data (contextual data: ERP, quality, MES, etc.), and manufacturing process information describing the synchronicity between the machines, and the rate of production flow. This semi-manual approach doesn’t take into account the more complex dynamic behavioral patterns of the machinery, or the contextual data relating to the manufacturing process at large. SCADA systems are used to monitor and control a plant or equipment in industries such as telecommunications, water and waste control, energy, oil and gas refining and transportation SCADA: A computer system for gathering and analyzing real-time data. While certain Facility Managers do perform predictive maintenance, this has traditionally been done using SCADA systems set up with human-coded thresholds, alert rules, and configurations.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |