Predictive Maintenance en Big Data
Predictive maintenance is a maintenance methodology based on the principle that a machine itself indicates the right time when the specific maintenance is needed. The objective of this methodology is to reduce costs and increase availability. Another important objective is to increase safety by preventing failures. This form of maintenance is made possible by the connection of smart sensors on machines. These sensors measure the condition of the machine and forward this information to the PM system. In everyday life, this kind of maintenance becomes more visible. Think of the cars we ride, which now tell us that oil should be added, that the thickness of the brake discs have a critical value, the air pressure of the tires is insufficient, etc.
Manufacturing company’s already apply predictive maintenance for many years. To check the condition of machinery, most companies have a program for using sensors to measure the vibration of bearings of rotating equipment, carry out analyzes of the oil used in machinery and measuring the temperature of critical electronic components. Based on the outcome of this data there will be decided whether maintenance is needed and what maintenance in that case will take be carried out. What most companies still are not doing is using the process data that is collected in the various shop floor systems (MES, SCADA). This process date contains “hidden” information about the status of the different machinery. However, this information is not easy to filter from this huge amount of data. What data can be used to analyze the situation of a given machine and how is this information found? This requires specialist teams with knowledge of the processes, of the possible failure modes that can occur and knowledge of statistical and analytical models.
The process data mentioned above are typical for Big Data. It constitutes a large data set, which cannot be processed with a regular database by the speed at which it enters, and the diversity of the data. The data also needs to be quickly processed to be able to collect the hidden data online. The arrival of the in-memory databases like SAP HANA now offer the ability to analyze Big Data. At this time, SAP is developing the tools to help the specialists to retrieve the necessary information from the processed data. For this, SAP has all kinds of statistical and transformation functions created to apply Big Data into SAP HANA system. In addition, the possibility should come in a smart way, either automatically detect patterns that indicate upcoming failures of a particular machine type. It then automatically processes the data by searching for these patterns and by recognition there can be warnings, so the security is increased and unnecessary downtime and costs are avoided.
Newitera is one of the partners of SAP to help introduce this new technique within the Netherlands. The objective this year is to apply the great opportunities that SAP HANA provides on the process data of an end customer. This offers you an unique opportunity in cooperation with Newitera (SAP partner in EAM) to benefit from a remarkable saving on maintenance costs while increasing your security.
For more information call Paul Hensgens: 06 11 75 10 38.
- Business Intelligence Solutions
- Development and Integration
- Management Consultancy
- Enterprise Asset Management EAM
- CRM Solutions
- Cloud Solutions
- Finance Solutions