Tools for Autonomic Resource Management in the Cognitive Continuum (ARMC3)

Important results from the project

A demonstrator that can show -the benefit (energy/cost/quality) of optimizing microservice placement in a realistic system, as well as the benefit of prediction to be proactive instead of reactive. -how machine learning can be used to find deviations in a complex realistic system. -how autonomous robots experience different latency when moving relative to a 5G antenna. A 3D visualization/simulation of an autonomous department store. Collaboration with both Umeå Universitet and Ericsson.

Expected long term effects

Contribute to achieving sustainability goals in the green transition. It will be easier to capture the interest of new customers New research results, which can form the basis for solutions that different companies can develop. More in-depth and extended collaboration with Umeå Universitet and Ericsson. We gain a deeper understanding of cloud solutions and microservices, which allows us to take broader responsibility for jobs similar to those we had in the project.

Approach and implementation

The structure: Project admin who was responsible for reporting. Project/Team leader who planned the work three weeks at a time, also handled external contacts. Three developers who were involved in the project the whole time and had a holistic view. 4 other developers were occasionally involved in the project and helped for a few months Changing experts caused too much overhead. As a result, we lost some speed and got started late with the integration of UmU´s optimizer. Good cooperation.