Friday 16th

from 16:15 pm to 17:00 pm

Room 19

-

Technical

PROTEUS will contribute to the EU Big Data area by addressing fundamental challenges related to the scalability and responsiveness of analytics capabilities. The requirements are defined by an industrial use case: the prediction of defects in hot strip mill (steelmaking industry); but the outcomes of PROTEUS are flexible and scalable to solve problems in many other domains.

In particular, the project will go beyond the current state of art with specific contributions: i) Real-time Machine Learning for advanced massive stream data analytics; ii) Real-time Hybrid Computation (batch data and data streams); iii) Real-time Interactive Visual Analytics for Big Data; iv) Enhancement of Apache Flink, the EU Big Data platform; and v) Industrial validation of the technology advances in a real industrial use case.

The PROTEUS impact vision is manifold:

• i) strategic, by reducing the gap and dependency from the US technology, empowering the EU Big Data industry through the enrichment of the EU platform Apache Flink;
• ii) economic, by fostering the development of new skills and new job positions and opportunities towards economic growth;
• iii) industrial, by demonstrating in a real industrial scenario how the technology advancement can help an industrial case to solve their needs, while at the same time demonstrating flexibility and scalability as to be applied in a number of other potential sectors and applications;
• iv) and technical, by advancing on hybrid and streaming analytic architectures, enabling scalable distributed ML strategies, and providing novel real-time interactive visualization methods economic growth;
• v) industrial, by demonstrating in a real industrial scenario how the technology advancement can help an industrial case to solve their needs, while at the same time demonstrating flexibility and scalability as to be applied in a number of other potential sectors and applications; and iv) technical, by advancing on hybrid and streaming analytic architectures, enabling scalable distributed ML strategies, and providing novel real-time interactive visualization methods.

During this talk, we will see what are the current limitations and gaps of scalable online machine learning and real-time visual analytics and we will overview the technical approach to address them that we are using in PROTEUS.