Research Themes

Osmotic Computing principles

The background and core principles of osmotic computing are introduced here. Osmotic computing is a new paradigm that can realize the fluid and elastic management of complex service compositions deployed over heterogeneous, dynamic, and evolving network and compute infrastructures spanning the device to cloud continuum. Our ongoing research focuses both on the technology enablers for OC, but also on what OC can do towards achieving edge intelligence, end-to-end automated service coordination, and continuous elasticity.

Osmotic Computing middleware and orchestration components

We have laid the design requirements for Osmotic Message-Oriented Middleware. Key aspects include how to model and implement osmotic pressure, how to provide the monitoring information necessary for it, and how to enable AI-driven management and orchestration of osmotic micro-elements (MELs).

Edge computing system support, tools, and algorithmic aspects

Osmotic Computing builds on top and helps to efficiently manage diverse federated cloud/fog/edge computing resources. We have extensively addressed two key aspects: (i) how to build and operate edge computing fabrics which include low-cost but also specialized edge compute units (such as those tailored to AI workloads), and (ii) how to handle near-optimally and at run-time service component placement (e.g., MELs) and workload distribution. We are also carrying out work in the evaluation of such systems: We are studying how to generate realistic edge computing topologies and scenarios and how to simulate edge/cloud computing systems with a view to pervasive IoT systems.

Enabling edge intelligence and continuous elasticity

Our work is in the space of Edge Intelligence, which represents the confluence of AI and Edge Computing. Osmotic Computing can play a key role in two directions: Pushing intelligence on the edge, by decentralized, autonomous decision making on-device and/or on edge compute resources, but also, and equally importantly, bringing intelligence for the edge, by enabling the smart orchestration of AI services and workloads across the device to cloud continuum, by dynamically, optimally and continuously (re)deploying and (re)configuring elastic AI pipelines, such as those involved in federated learning or inference by means of Distributed Deep Neural Networks.

Osmotic computing application scenarios and case studies

Osmotic Computing as a generic model for distributed computation has been studied in various application contexts, including e-Health in general, and body-area networking in particular, and IoT workflows.

Copyright Policy on Published Papers
Several papers are available for download. By following these links you agree to respect the copyrights of the papers.

The papers obtained from this Web page are included by the contributing authors as a means to ensure timely dissemination of scholarly and technical work on a non-commercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author’s copyright. These works may not be reposted without the explicit permission of the copyright holder.

%d bloggers like this: