Periodic Reporting for period 1 - ASSIST-IoT (Architecture for Scalable, Self-*, human-centric, Intelligent, Secure, and Tactile next generation IoT)
Período documentado: 2020-11-01 hasta 2022-04-30
ASSIST-IoT is a Research and Innovation Action that aims at creating a novel NG-IoT architecture addressing scalability and flexibility of data processing and analytics, by implementing a decentralized and secure solution, supported by smart components (i.e. devices, edge nodes, networks, applications and services), allowing execution of context-aware applications with new interaction interfaces (e.g. AR/VR/MR). The blueprint architecture to be delivered by the project will cover the emergent and upcoming needs of next generation IoT deployments. Apart from establishing the basic technology for such solutions, ASSIST-IoT will indicate how to incorporate new “modules” (enablers) to achieve diverse NGIoT innovations upon. The solution will integrate AI-based functions transferring intelligence closer to the edge (data sources), including devices. The reference architecture will be validated in three realistic pilots, remarking the use-case orientation of the project results: (i) port automation; (ii) smart safety of workers, and (iii) cohesive vehicle monitoring and diagnostics. One of the main objectives is to build this architecture following a human-centric approach. This means, in the context of IoT, that the humans using / being part of the IoT deployment are at its very centre, regarding usability, data protection, rights and freedom preservation, and the computing itself. For achieving so, ASSIST-IoT will incorporate in its architecture a suite of technologies (“enablers”) like DLT, trust, AI and cybersecurity.
The specific objectives are as follows:
-Obj.1: Design, implementation and validation of an NGIoT Reference Architecture
-Obj.2: Definition and implementation of distributed smart networking components
-Obj.3: Definition and implementation of decentralized security and privacy exploiting DLT
-Obj.4: Definition and implementation of smart distributed AI enablers
-Obj.5: Definition and implementation of human-centric tools and interfaces
-Obj.6: Definition, deployment and evaluation of real-life pilots.
-Obj.7: Establishment of an innovative cooperation and business framework
- Obj.8: Impact creation: Showcasing ASSIST-IoT and Disrupting the current market
The specific objectives are as follows:
-Obj.1: Design, implementation and validation of an NGIoT Reference Architecture
-Obj.2: Definition and implementation of distributed smart networking components
-Obj.3: Definition and implementation of decentralized security and privacy exploiting DLT
-Obj.4: Definition and implementation of smart distributed AI enablers
-Obj.5: Definition and implementation of human-centric tools and interfaces
-Obj.6: Definition, deployment and evaluation of real-life pilots.
-Obj.7: Establishment of an innovative cooperation and business framework
- Obj.8: Impact creation: Showcasing ASSIST-IoT and Disrupting the current market
The first semester of the project was focused on setting the foundations towards one of the main research objectives of the work programme: “providing a reference implementation in terms of dynamically configured infrastructure”. All the work performed in T3.5 WP4 and WP5 in the period M1-M6 was oriented towards this approach, describing the design principles of ASSIST-IoT architecture as a strong basis (with innovative technologies) that grows via the introduction of specific enablers in each deployment (i.e. pilots in ASSIST-IoT). In addition, all the technical topics relevant for the programme (interoperability, IoT devices, contextual IoT, network…) have been being specifically addressed in the project via dedicated tasks.
After M6, the project entered a steadier phase that has lasted till the very date of delivery of this document (around M18 – April 2022). This period has witnessed a substantial amount of work in several fronts: (i) technical; advancing towards the final formalisation of the architecture (will be completed by M21), refining an intermediate version, and evolving its inner components (the “enablers”), that have been fully described, carefully bounded, designed, planned and (initially) developed – this has included selection of technologies and (in most of them), showcasing MVP versions. (ii) In terms of pilots, initial scope has been enhanced towards a closed definition of extent, via business scenarios, use cases and requirements (in the sense of WP3) and trials (WP7); concrete hardware procurement, software development and overall integration plans have been agreed and scheduled, and first tangible advances are being noticed (see deliverable D7.2). Furthermore, first glance upon impact KPIs has been performed via the planning of Evaluation and Assessment activities in ASSIST-IoT. Details on how these have been defined can be found at Section 2.1 of the Technical Report.
At the same time, impact outreach activities have been uninterruptedly tackled during the whole M1-M18 period. Communication via social media and exposure through website and other channels is continuously address by specialised teams as well as by partners individually. This also applies to scientific dissemination events, which little by little are turning again towards face-to-face fashion, allowing better spread of ASSIST-IoT achievements. Finally, the first round of Open Calls has been initiated and completed, engaging 37 candidates (SMEs, RTOs and Universities included), redounding in the joining of 7 external individual entities that leverage ASSIST-IoT funding to deliver innovative, single projects bringing added value to one of our pilots while validating (the entirety or part of) our architecture.
After M6, the project entered a steadier phase that has lasted till the very date of delivery of this document (around M18 – April 2022). This period has witnessed a substantial amount of work in several fronts: (i) technical; advancing towards the final formalisation of the architecture (will be completed by M21), refining an intermediate version, and evolving its inner components (the “enablers”), that have been fully described, carefully bounded, designed, planned and (initially) developed – this has included selection of technologies and (in most of them), showcasing MVP versions. (ii) In terms of pilots, initial scope has been enhanced towards a closed definition of extent, via business scenarios, use cases and requirements (in the sense of WP3) and trials (WP7); concrete hardware procurement, software development and overall integration plans have been agreed and scheduled, and first tangible advances are being noticed (see deliverable D7.2). Furthermore, first glance upon impact KPIs has been performed via the planning of Evaluation and Assessment activities in ASSIST-IoT. Details on how these have been defined can be found at Section 2.1 of the Technical Report.
At the same time, impact outreach activities have been uninterruptedly tackled during the whole M1-M18 period. Communication via social media and exposure through website and other channels is continuously address by specialised teams as well as by partners individually. This also applies to scientific dissemination events, which little by little are turning again towards face-to-face fashion, allowing better spread of ASSIST-IoT achievements. Finally, the first round of Open Calls has been initiated and completed, engaging 37 candidates (SMEs, RTOs and Universities included), redounding in the joining of 7 external individual entities that leverage ASSIST-IoT funding to deliver innovative, single projects bringing added value to one of our pilots while validating (the entirety or part of) our architecture.
ASSIST-IoT reflects European need to create and offer technology enabled solutions facilitating execution of activities from different verticals, improving operational performance, and providing security to data and people involved in those activities. To sum up, ASSIST-IoT will help to advance the state of the art in IoT architectures and enablers.
The advances on the state of the art will be noticeable from various perspectives:
• Add analytics and AI capabilities (e.g. to support federated learning) to device and edge plane nodes. Implementation of containerisation capabilities in edge nodes.
• Adapt the orchestrator to NG-IoT environments, to work with constrained-resource container orchestrators such as K8s or K3s.
• Provide NFVs and AI models that improve the network performance for NG-IoT use cases.
• Deliver ontologies and semantic enablers for fully-fledged support of semantic data.
• Implement a set of data processing functionalities using DLT, intelligent edge data brokerage, and high-availability data storage and management, all governed by sensitive data processing rules.
• Support innovative interaction mechanisms like AR/MR interfaces
• Development of a new smart device (wearable-like) that will be demonstrated in the context of a construction site.
• Development of a new UI-AR/MR tool that will be demonstrated in the context of a maritime port.
• To allow a NGIoT deployment to be self-healer, self-protected, self-aware, self-organised and self-configured as a system in an automated way, ranging from far edge devices to the cloud.
• Support the federation of learning to solve problems with AI under a collaborative, distributed fashion, sharing intelligence (models, results) among different nodes of an IoT deployment.
• Achieve co-living and co-operation of cybersecurity tools (authentication, identity management, threat detection) with modern DLT tools and concepts (Smart Contracts, logging and auditing), in all nodes of the IoT deployment, all of that harmonised with data semantics to achieve sovereignty.
• Making the NGIoT deployment human-centric via a single window-fashion framework to manage all aspects of the system (deployment, configuration, representation, results, users, network, devices…)
The advances on the state of the art will be noticeable from various perspectives:
• Add analytics and AI capabilities (e.g. to support federated learning) to device and edge plane nodes. Implementation of containerisation capabilities in edge nodes.
• Adapt the orchestrator to NG-IoT environments, to work with constrained-resource container orchestrators such as K8s or K3s.
• Provide NFVs and AI models that improve the network performance for NG-IoT use cases.
• Deliver ontologies and semantic enablers for fully-fledged support of semantic data.
• Implement a set of data processing functionalities using DLT, intelligent edge data brokerage, and high-availability data storage and management, all governed by sensitive data processing rules.
• Support innovative interaction mechanisms like AR/MR interfaces
• Development of a new smart device (wearable-like) that will be demonstrated in the context of a construction site.
• Development of a new UI-AR/MR tool that will be demonstrated in the context of a maritime port.
• To allow a NGIoT deployment to be self-healer, self-protected, self-aware, self-organised and self-configured as a system in an automated way, ranging from far edge devices to the cloud.
• Support the federation of learning to solve problems with AI under a collaborative, distributed fashion, sharing intelligence (models, results) among different nodes of an IoT deployment.
• Achieve co-living and co-operation of cybersecurity tools (authentication, identity management, threat detection) with modern DLT tools and concepts (Smart Contracts, logging and auditing), in all nodes of the IoT deployment, all of that harmonised with data semantics to achieve sovereignty.
• Making the NGIoT deployment human-centric via a single window-fashion framework to manage all aspects of the system (deployment, configuration, representation, results, users, network, devices…)