USU develops new data science application for Industry 4.0
06 March 2018
A new graphical development environment for data science in mechanical and plant engineering is now available in the shape of KatanaFlow.
The powerful, user-friendly web application supports engineers in analysing and processing industrial data. Ready-made modules and components can be dragged and dropped from extensive machine learning and data mining libraries, meaning even non-data scientists at departments are able to create models that deliver insightful results. KatanaFlow specifically provides support in data preprocessing, cleansing and correlation, as well as in creating time series analyses, recognising patterns and anomalies, and many other evaluations. The findings obtained from that also form the basis for new digital services.
Data scientists spend about 80 percent of their working time on preparing and processing complex machine data. Graphical programming of data evaluations that can be compiled interactively from ready-made algorithm modules created in-house slashes the time and effort involved in complex data analyses. All users benefit from being able to create evaluations in a clear way by means of self service, as well as from the flexible collaboration options the cloud application offers.
A demo version of KatanaFlow is available at: https://katana.usu.de/en/
Katana has been a USU division since 2017 and specialises in big data analytics, a key field in digitisation. Successful industry projects show the market potential of data-driven digital service offerings.