WASP WARA – Operational Data Research Arena

Project description

WARA-Ops aims to actively enable and encourage a significant increase in research collaborations between Swedish industry and members of WASP, The Wallenberg AI, Autonomous Systems and Software Programme, by providing a unique national resource. WARA-Ops is hosting community events (seminars, workshops, talks) and developing a Cloud-based data storage and processing portal that allows companies and researchers to store and process datasets in an easy-to-use and secure manner.

Key areas

Although WARA-Ops supports all kinds of data-driven operational research, we have a current focus on two key areas:

• Anomaly detection

  We collect and provide datasets and associated research challenges, which may include intrusion detection and denial-of-service (DoS), attacks. Examples of data include system and software log files and access request lists.

• Predictive analytics using time-series data

  We collect and provide datasets that feature time-series data, and host research challenges related to this – e.g. how to accurately predict future behaviour using as little historical data as possible, how to efficiently detect concept drift, etc. Examples of such data are system metrics such as CPU utilisations over time within a data centre, temperatures over time within rooms in a building, etc.

Wide range of application areas

The above research results are applicable in a wide range of application areas, including most online services, communication systems and automation services as well as in energy management, financial and banking systems, online retail and cloud/edge data centres. Application areas also include software building systems and CI/CD pipelines.

In partnership with Ericsson Research

With a strong focus on data storage and cataloguing, combined with the provision of a scalable CPU-based environment for interactive data querying, processing and visualisation, WARA-Ops is a natural complement to the Berzelius service, which focuses on GPU-intensive AI/ML workloads. The project is a continuation of WARA-Common and is professionally managed by Ericsson Research, significantly expanding the current service. WARA-Ops will also gain access to a live 5G network and a full-scale cloud service, as well as access to the monitoring data generated.