Periodic Reporting for period 1 - OASEES (Open Autonomous programmable cloud appS & smart EdgE Sensors)
Período documentado: 2023-01-01 hasta 2024-06-30
OASEES aims to create an open, decentralized, intelligent, programmable edge framework for Swarm architectures and applications, leveraging the Decentralized Autonomous Organization (DAO) paradigm and integrating Human-in-the-Loop (HITL) processes for efficient decision making.
The OASEES vision is to provide the open tools and secure environments for swarm programming and orchestration for numerous fields, in a completely decentralized manner. An important aspect in this process is identification and identity management, in which OASEES targets the implementation of a portable and privacy preserving ID federation system, for edge devices and services, with full compliance and compatibility to GAIA-X federation and IDSA trust directives and specifications. This situation solidifies the need for an integrated enabler framework tailored to the edge’s extreme data processing demands, using different edge accelerators, i.e. GPU, NPU, SNN and Quantum.
The OASEES project aims to directly address these challenges, by delivering and promoting a European, fully open-source, decentralized and secure Swarm programmability framework for edge devices and leveraging various AI/ML accelerators (FPGAs, SNNs, Quantum), while supporting a privacy preserving Object ID federation process
Objective 1: Design a decentralized, agile and secure architecture for collaborative smart nodes at the edge, supporting heterogeneous device communication, backed by the Decentralized Autonomous Organization (DAO) paradigm integration.
Objective 2: Build a secure, trustworthy and decentralized edge ecosystem with native device support by integrating Self Sovereign Identity (SSI) for a portable digital identity that does not depend on any centralized authority. The OASEES decentralized device identity will be a new class of identifier that fulfils all four requirements: persistence, global resolvability, cryptographic verifiability, and decentralization.
Objective 3: Build rapid development kits (RDKs) for an open programmable framework across different smart edge nodes, while incorporating efficient cloud-to-edge continuum intelligence across heterogeneous target environments.
Objective 4: Demonstrate the framework and programmability toolkit in a set of different vertical use cases and evaluate the benefits across different sectors.
Objective 5: Maximize the impact of the OASEES results via extensive communication, scientific dissemination, and exploitation activities. Foster the creation of an open-source community around the OASEES solution, engaging a diverse set of stakeholders.
The OASEES vision is to provide the open tools and secure environments for swarm programming and orchestration for numerous fields, in a completely decentralized manner. An important aspect in this process is identification and identity management, in which OASEES targets the implementation of a portable and privacy preserving ID federation system, for edge devices and services, with full compliance and compatibility to GAIA-X federation and IDSA trust directives and specifications. This situation solidifies the need for an integrated enabler framework tailored to the edge’s extreme data processing demands, using different edge accelerators, i.e. GPU, NPU, SNN and Quantum.
The OASEES project aims to directly address these challenges, by delivering and promoting a European, fully open-source, decentralized and secure Swarm programmability framework for edge devices and leveraging various AI/ML accelerators (FPGAs, SNNs, Quantum), while supporting a privacy preserving Object ID federation process
Objective 1: Design a decentralized, agile and secure architecture for collaborative smart nodes at the edge, supporting heterogeneous device communication, backed by the Decentralized Autonomous Organization (DAO) paradigm integration.
Objective 2: Build a secure, trustworthy and decentralized edge ecosystem with native device support by integrating Self Sovereign Identity (SSI) for a portable digital identity that does not depend on any centralized authority. The OASEES decentralized device identity will be a new class of identifier that fulfils all four requirements: persistence, global resolvability, cryptographic verifiability, and decentralization.
Objective 3: Build rapid development kits (RDKs) for an open programmable framework across different smart edge nodes, while incorporating efficient cloud-to-edge continuum intelligence across heterogeneous target environments.
Objective 4: Demonstrate the framework and programmability toolkit in a set of different vertical use cases and evaluate the benefits across different sectors.
Objective 5: Maximize the impact of the OASEES results via extensive communication, scientific dissemination, and exploitation activities. Foster the creation of an open-source community around the OASEES solution, engaging a diverse set of stakeholders.
From M1 to M18, OASEES focused on developing a detailed technical architecture ensuring compatibility and interoperability of components. The project integrated advanced technologies such as edge computing, federated learning, and decentralized identifiers. Innovation management involved incorporating new technological developments and leveraging Self-Sovereign Identity (SSI) technologies for enhanced security and privacy. The project developed formal specifications for DApps, DAOs, and smart contracts through use case diagrams and sequential diagrams.
The project focused on developing interfaces and endpoints for data federation and accelerator integration. Detailed sections for each layer of the OASEES stack were implemented, with particular emphasis on data federation relevant to GAIA-X and accelerator integration covering a wide scope of enablers from quantum processing units to GPUs and SNNs.
In the frame of OASEES, the Quantum DevKit - Eclipse Qrisp was co-developed and released. Firstly, as a standalone release and then under the umbrella of the Eclipse foundation.
Eclipse Qrisp is the first European Quantum Devkit, which enables quantum programming in a user friendly manner, requiring only the knowledge of Python. It has been fully integrated in the OASEES SDK, and has been applied in a number of PoCs regarding swarm optimization problems.
The project provided the testbed for validating the MLOps pipeline of the SDK and conducted federated ML tests. Modules and interfaces for the connectivity continuum in the OASEES SDK were developed, enabling automatic workload assignment to specific edge devices . Focused on developing and testing the connectivity continuum for the OASEES SDK. This included implementing platform-specific modules for different edge devices, conducting 5G-RedCap protocol tests, and ensuring efficient energy performance and network connectivity.
The project focused on developing interfaces and endpoints for data federation and accelerator integration. Detailed sections for each layer of the OASEES stack were implemented, with particular emphasis on data federation relevant to GAIA-X and accelerator integration covering a wide scope of enablers from quantum processing units to GPUs and SNNs.
In the frame of OASEES, the Quantum DevKit - Eclipse Qrisp was co-developed and released. Firstly, as a standalone release and then under the umbrella of the Eclipse foundation.
Eclipse Qrisp is the first European Quantum Devkit, which enables quantum programming in a user friendly manner, requiring only the knowledge of Python. It has been fully integrated in the OASEES SDK, and has been applied in a number of PoCs regarding swarm optimization problems.
The project provided the testbed for validating the MLOps pipeline of the SDK and conducted federated ML tests. Modules and interfaces for the connectivity continuum in the OASEES SDK were developed, enabling automatic workload assignment to specific edge devices . Focused on developing and testing the connectivity continuum for the OASEES SDK. This included implementing platform-specific modules for different edge devices, conducting 5G-RedCap protocol tests, and ensuring efficient energy performance and network connectivity.
From M1 to M18, the OASEES project has made significant progress in building a decentralized edge computing system. The key technical achievements include:
Created a system for running large, distributed machine learning (ML) tasks.
Analytics Proof-of-Concept (PoC) Solutions:
Developed solutions for tasks like image recognition, detecting unusual events, and predicting future trends.
Integrated advanced AI technologies like Spiking Neural Networks (SNN) and Quantum processors.
Implemented secure communication protocols and decentralized identity management using Self-Sovereign Identity (SSI) and Decentralized Autonomous Organizations (DAOs).
Developed AI tools to check smart contracts for vulnerabilities and conducted hardware checks for use-case equipment.
Using SSI and blockchain technology ensures secure and decentralized management of identities.
Privacy-preserving techniques and decentralized data handling reduce the risks associated with centralized data collection.
Innovation and Technological Advancement:
Developing advanced AI technologies, including SNN and quantum computing, keeps OASEES at the cutting edge of technology.
The wide range of services supports further research and development in edge computing.
Ready-to-use models and a robust SDK make it easier to bring OASEES solutions to the market.
The secure and efficient infrastructure makes OASEES solutions attractive to businesses, driving adoption and growth.
Key Needs for Further Success
Conduct large-scale demonstrations and pilot projects to showcase OASEES capabilities in real-world scenarios.
Collaborate with industry partners to test and validate solutions in various environments.
Created a system for running large, distributed machine learning (ML) tasks.
Analytics Proof-of-Concept (PoC) Solutions:
Developed solutions for tasks like image recognition, detecting unusual events, and predicting future trends.
Integrated advanced AI technologies like Spiking Neural Networks (SNN) and Quantum processors.
Implemented secure communication protocols and decentralized identity management using Self-Sovereign Identity (SSI) and Decentralized Autonomous Organizations (DAOs).
Developed AI tools to check smart contracts for vulnerabilities and conducted hardware checks for use-case equipment.
Using SSI and blockchain technology ensures secure and decentralized management of identities.
Privacy-preserving techniques and decentralized data handling reduce the risks associated with centralized data collection.
Innovation and Technological Advancement:
Developing advanced AI technologies, including SNN and quantum computing, keeps OASEES at the cutting edge of technology.
The wide range of services supports further research and development in edge computing.
Ready-to-use models and a robust SDK make it easier to bring OASEES solutions to the market.
The secure and efficient infrastructure makes OASEES solutions attractive to businesses, driving adoption and growth.
Key Needs for Further Success
Conduct large-scale demonstrations and pilot projects to showcase OASEES capabilities in real-world scenarios.
Collaborate with industry partners to test and validate solutions in various environments.