BB24.D Big Data Analytics

From its-wiki.no

Revision as of 17:45, 13 July 2017 by Josef.Noll (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
Title Big Data Analytics
Page Title BB24.D Big Data Analytics
Technology Line Distributed Cloud Integration
Lead partner ITI
Leader Antonio Lagarda
Contributors ITI
Related to Use Cases SCOTT:WP7, SCOTT:WP9, SCOTT:WP11, SCOTT:WP15
Description This building block will develop a Context-aware and reasoning platform based on Big Data cloud architecture able to store and process large amounts of data collected by sensors. The platform will apply and adapt the suitable Big Data services for data analytics, essentially based on machine learning techniques, to predict future sensors measurements, to detect anomalies on current sensors measurements, and to optimize processes or configurations of devices.
Main output This building block will be a software platform able to ingest data collected by sensors, analyse them and provide the analyses results to external applications. This platform will be composed of different components, integrated and modified to achieve the expected result in the destination scenarios and use cases. It will integrate scalable, robust and reliable technologies to store and process large amounts of data. Machine learning techniques will be applied for future values prediction, data anomalies detection, and optimization. The prototype will be tested and demonstrated in some use cases.
BB category SW component
Baseline This Building Block will start from the Context-aware and reasoning module developed in DEWI, which included the following main components: communications interface (RESTful API, JSON); Big Data storage (HDFS Hive); distributed computing (Spark); Big Data Analytics (Spark MLLib, H2O); alerts engine (WSO2 Complex Event Processor); and front-end (AngularJS, NodeJS, SailsJS). DEWI applied machine learning techniques based on deep learning to predict future measurements and to detect anomalies. SCOTT will adapt the platform’s communication interfaces; data preparation modules; and data analytics techniques according to the new use cases needs. In addition, SCOTT will extend this platform with further data analytics techniques for optimization purposes.
Current TRL 4. In this Building Block, different SW components designed in DEWI will be integrated and adapted to new use cases, and extended with new functionality.
Target TRL 6. A prototype Big Data Analytics platform will be implemented and demonstrated in selected use cases.