Article by Anastasios Zafeiropoulos
Technical Manager of the NEPHELE Horizon Europe Project, Postdoc Researcher at the National
Technical University of Athens
The next generation Internet of Things (IoT) and Edge Computing technologies are evolving at a rapid pace, aiming to optimally serve distributed applications with strict Quality of Service (QoS), security and data privacy requirements that can be deployed across resources in the computing continuum. This evolution moves in parallel with the increase in the heterogeneity of the IoT technologies in terms of the production of different types of intelligent IoT devices, the support of diverse communication protocols, and the conceptualization of various information models for semantically representing entities in the IoT world. These trends make inherent the need for novel architectural approaches, able to support by design a full convergence and integration among existing and evolving IoT and edge computing technologies.
Two main challenges arise. The first challenge regards the need for convergence of IoT technologies based on novel architectural approaches, able to guarantee continuous and seamless openness and interoperability of the plethora of existing and emerging solutions. The second challenge regards the need for the provision of open, modular and extensible meta-orchestration environments, able to manage the lifecycle of distributed applications across resources in the IoT, edge and cloud part of the computing continuum. Various enabling technologies can be considered to tackle these challenges.
Virtualization of IoT devices and functions through the provision of virtual counterparts of the IoT devices at the edge part of the infrastructure helps to tackle existing IoT interoperability issues. Such an approach is envisioned by ongoing activities in standardization working groups (e.g., the W3C Web of Things (WoT) working group) as well as within the development of an open-source software stack in the NEPHELE project (called as Virtual Object Stack). Through a set of abstractions and software implementations, IoT solutions’ providers and application developers can provide virtualized instances of the IoT infrastructure, while develop a set of functions that can be executed in a device-independent way. Interoperability can be achieved in both protocol and semantic level, through the support of various communication protocols and translators among semantic models in the software stack. The concept of Digital Twins (DTs) is also considered, since a DT can be easily developed by taking advantage of the provided virtualized IoT infrastructure.
Moving one step further, the development of distributed applications that integrate IoT functions, DTs, and inherently support the interaction with IoT devices is facilitated. IoT functions can be part of a distributed application graph that is manageable by emerging meta-orchestration platforms. In this
way, convergence of IoT with edge and cloud computing technologies is strengthened, since the developed applications can be deployed over resources that span across the computing continuum. However, the development of such solutions has to be accompanied by the provision of well-formulated abstractions for the management of heterogeneous resources, enabling the translation of high-level objectives to deployment and runtime management plans in an agnostic way for application developers.
Modularity and openness should be considered by design in the development of such solutions, enabling their adoption and extension by both the research and industrial communities.
For further information on the Unlock-CEI project or to share success stories, please subscribe to our newsletter.