Technology
Technical aspect
Data Mining, also known as Knowledge Discovery in Databases (KDD), is an information extraction activity aiming at discovering new knowledge and facts from large databases. Data Mining uses a broad range of tools from statistics, automatic learning, pattern recognition, database technologies, visualization and artificial intelligence. This mix of technologies is the core of PEPITo platform, the Data Mining environment developed by PEPITe.
Methodical aspect
In addition, Data Mining needs to be regarded as a methodical process. A rigorous methodology will allow our customers to maximize the return on investments (ROI) of a Data Mining project (risk minimization, meet scheduled deadlines, clear definition of objectives and responsibilities,...).
Practically, PEPITe offers services copied from the CRISP-DM approach (CRoss Industry Standard Process for Data Mining). This standard has been developed by a consortium of experienced industrials in order to clearly define the successive steps of a Data Mining project.
Industrial applications of Data Mining
Today, the complexity of modern manufacturing processes in a highly competitive environment forces the manufacturers to invest massively in automation and monitoring systems. Unfortunately, these new installations are generating so large data flows that these sources of valuable and hidden improvements are more and more underused.
On the other side, Six Sigma, ISO9000 2000 quality management standards, etc. are clearly pushing up this idea of continuously monitoring your system to draw objectively improvement actions from this information cycle. In this context, Data Mining plays an obvious role in the continuous improvement of manufacturing processes.
The information lifecycle, a driver of the continuous improvement
|
|
Manufacturing systems monitoring: sensors, PLCs, DCS, and SCADA systems allow the operators to monitor and control the manufacturing process in real time.
|
|
|
Data organization and storing: the whole measurements and actions (manual and automatic) made on the process are recorded and stored in historians that represent huge memories of the factory.
|
|
|
Historical data analysis: experts are digging into the data off-line to detect flaws and improvement actions. This is also a unique opportunity to learn more quickly about the process and to detect hidden and complex relationships between all parameters. Given the increasing amount of these archives, the Data Mining solutions developed by PEPITe are more than welcomed to maximize the benefits of this data analysis task.
|
|
|
Improvement actions: at the end of this life cycle, the knowledge synthesized by the Data Mining analysis is used by operators to bring improvements to their processes. This knowledge might also be recorded in a knowledge base used by an artificial expert system.
|
By an optimal use of available data, PEPITe makes it possible to close efficiently the information cycle and to increase dramatically the revenues expected by monitoring investments.
|
News