HORIZON 2020
INTEND
Project name and Acronym
INTEND – Intent-based Data Operation in the Cognitive Computing Continuum
Project Code
101135576
Call
HORIZON-CL4-2023-DATA-01-04
Financing Program
Horizon Europe
Promotor/ Coordinator
Sintef AS
Consortia Partners
Sintef AS | Sapienza Università di Roma | Technische Universität Wien | Intel Deutschland GMBH | DELL | Nextworks | Onlim GMBH | Ericsson | CS-Group Romania | MOG Technologies | FILL GMBH | Telenor ASA | GATE Institute Sofia University | Seoul National University | AiM Future, Inc. | Hanyang University
Duration
36 months
Total Eligible Costs (€)
346 250.00
EU Grant Amount (€)
346 250.00
Main objective
INTEND aims to bring human-like intelligence into the cognitive continuum, to achieve the novel concept of intent-based data operation, which is capable of: learning how to manage and adapt diverse cloud/edge resources for more efficient data processing, strategic decision making across the decentralized continuum for end-to-end data security and sustainability, and human-friendly interaction with data stakeholders. The consortium will deliver 11 software tools, validated on 5 domains: video streaming, machine data platform, 5G data infrastructure, urban data space, and robotic AI.
Objectives
Objective 1: Define the intent-based data operation concept, requirement, architecture, and dataset.
Objective 2: Achieve intelligent management of computing, storage, network, and neural processing resources based on continual learning, to better exploit diverse and unconventional hardware on edge and cloud.
Objective 3: Enable strategic and federated decision making covering multiple AIs from different perspectives and places across the continuum, to achieve end-to-end data security and sustainability.
Objective 4: Achieve human-centric data operation, via a novel nature language interface for data stakeholders to efficiently express their data operation intents and understand and trust the AI-made decisions.
Objective 5: Build an integrated INTEND toolbox for intent-based data operation, together with an open platform for the extension and interoperability with new hardware, operation tools, and AI models.
Objective 6: Test and validate the technical advancements from the INTEND toolbox through five unique use cases applied to media, manufacturing, telecom, smart cities and logistics sectors.
Objective 7: Foster the open edge ecosystem in Europe via the uptake of INTEND results, the research and industry community building, and the international collaboration.
Work Packages
Work Package 1: Requirement, architecture and use case definition.
Work Package 2: Intelligent resource management based on continual learning.
Work Package 3: Federated and strategic decision making for cognitive continuum.
Work Package 4: Chat-based stakeholder interface for intent-based data operation.
Work Package 5: Consolidation of research and development of prototype tools.
Work Package 6: Integration and open platform for intent-based data operation.
Work Package 7: UC1: Intent-aware green video streaming.
Work Package 8: UC2: Chat-based operation of machine data analytics platform.
Work Package 9: UC3: Intent-based 5G infrastructure for data and AI pipelines.
Work Package 10: UC4: Intent-based platform for operational urban data space.
Work Package 11: UC5: Adaptive federated learning on robotic systems.
Work Package 12: Dissemination and Exploitation 1.
Work Package 13: Dissemination and Exploitation 2.
Work Package 14: Project Management 1.
Work Package 15: Project Management 2.