Authors:
Vincenza Carchiolo
1
;
Alessandro Longheu
1
;
Giuseppe Saccullo
2
;
Simona Maria Sau
2
and
Renato Sortino
1
Affiliations:
1
University of Catania, Italy
;
2
STMicroelectronics, Catania, Italy
Keyword(s):
Log Mining, Manufacturing Workflow, Big Data, Real-time Analysis, Elasticsearch, Data Visualization.
Abstract:
Today’s complexity in ICT services, consisting of several interacting applications, requires strict control over log files to detect what exceptions/errors occurred and how they could be fixed. The current scenario is harder and harder due to the volume, velocity, and variety of (big) data within log files, therefore an approach to assist developers and facilitate their work is needed. In this paper an industrial application of such log analysis is presented, in particular, we consider the manufacturing of Integrated Circuits (ICs), i.e. a set of physical and chemical processes performed by production machines onto silicon slices. We present a widely used set of open-source tools that join together a platform to allow logs mining to assess manufacturing workflow processes. We show that the proposed architecture helps in discovering and removing anomalies and slowdown in ICs production.