CloudSkin: Adaptive virtualization for AI-enabled Cloud-edge Continuum

CLOUDSKIN aims to design a cognitive cloud continuum platform to fully exploit the available Cloud-edge heterogeneous resources, finding the “sweet spot” between the cloud and the edge, and smartly adapting to changes in application behavior via AI. To facilitate automatic deployment, mobility and security of services, CloudSkin will build an innovative universal container-like execution abstraction based on WebAssembly that allows the seamless and trustworthy execution of (legacy) applications across the Cloud-edge continuum.

• Smart management for the Cloud-edge continuum: Apply AI methods to transparently orchestrate Cloud-edge resources, in cooperation with the application execution framework and continuum infrastructure.
• Virtual execution for the Cloud-edge continuum: Developing a universal execution abstraction paired with trusted computing, we called it “Cloud-edge cells”, to enable the trusted execution of applications across the continuum.
• Infrastructure support for the Cloud-edge continuum: Design an infrastructure to deliver bare metal resource performance to storage, despite virtualization and dynamic reallocation.

Role of Cellnex:  Cellnex provides the Castellolí small-scale test site is located 55 km north-west of Barcelona and is part of the Parcmotor Circuit, illustrated in the figure, a test track rented by OEMs (Original Equipment Manufacturers) for stress tests of vehicles. It features one private 5G SA networks operated by Cellnex, and its premises host both the test 5G network and the datacenter where most of the services are running. Castellolí small-scale test site provides a reliable testing environment that enables partners to refine and improve the services before deployment in real-world scenarios.

In the Cloudskin project, Cellnex provides all the EDGE infrastructure and the circuit as a test site for the use cases.

Use Cases:

• Edge orchestration and video analytics: Orchestration of edge apps with matching cloud performance and the creation of AI video-analytics.
• Surgery: Real-time edge video analytics with dynamic resource allocation and Private Deep & Federated Learning at the edge.
• Metabolomics: Edge/on-premise batch analytics and reduction of cloud offloading for the METASPACE metabolite annotation platform.
• Agriculture IoT: Dynamic cloud offloading to match detail level and creation of an IoT-based agriculture data space.

Partners:

Funding from: