The exponential growth of intelligent sensors and devices is generating an unprecedent amount of data as well as new approaches to consume them. Around 75% of data is expected to be produced at the edge by 2025, with only 25% still produced within the data center. This is reshaping IT architectures, as increasingly powerful processing and machine learning inference capabilities are required at the edge of the networks to enable next generation, transformative AI and IoT applications. Edge computing keeps a bridge to datacenters and cloud. It offers processed data to be send for further analytics in cloud and on-premises applications. As well, a cloud-to-edge infrastructure, allows to run cloud native application at the edge to deliver analytics in real time in case of critical needs.
Current computing systems are complex, as they span across the edge-cloud continuum, creating a landscape where core clouds should jointly operate with highly distributed edge systems. Our mission focuses on meeting these challenges to deliver a novel AI-assisted, programmable, highly automated and adaptive Cloud-Edge continuum framework for cognitive resources and applications management, leveraging green, elasticity and data processing demands and attributes as well as a proper modelling and profiling of the overall context to feed the AI-assisted decision-making processes. On the other hand, the media team has the mission to work towards the development of innovative assets and tools focusing on immersive media management and distribution, multiplatform eXtended Reality Interaction.
The immersive media & edge technology research unit encompasses two different and complementary research lines: The new media and interactive technologies (NMIT) research line and the edge and software technology (EST) research line.
The expertise of our team makes it possible to cover areas including Immersive Media and Extended Reality, Edge and Cloud Service Management, Service Compliant Policy Management
The team is working on the following key research lines: