Data Mining and Big Data techniques for predictive plant maintenance


Thu 21 May 2015

Event description

The processing of large amounts of data produced by the sensor systems applied to production plants is a complex activity due not only to the volume of data involved but also to their variety, the speed with which they vary, their veracity. These are analysis activities that are gaining increasing attention in advanced and structured industrial environments, often in relation to the term “Big Data”.

Manufacturing companies possess a great wealth of information, deriving from maintenance interventions, from sensor systems applied to machines, which crossed with other exogenous data to the company can provide useful indications to improve the maintenance service itself in terms of cost reduction and inconvenience for customers.

The proposed meeting aims to investigate the possibilities offered by Data Mining and Big Data techniques applied to integrated data coming from multiple internal and external sources for a more reactive and predictive maintenance of the plants.
Given that for many companies that of Big Data is still only a matter of perspective, in this phase it can be very interesting to understand:

  • What data is collected (source, structure, type)
  • How they are collected
  • How they are managed (saved, filtered, archived)
  • How (and where) they are processed and analyzed
  • What can be achieved by systematic management and analysis of heterogeneous data

Get in touch