Create a trustworthy edge AI ecosystem including an AI architecture, components, orchestration techniques, development tools and a community enabling the real-time collaboration of heterogeneous edge devices, while maintaining the highest levels of security, scalability, sustainability and robustness.
Having assembled some of Europe’s best partners from industry, research, and academia, EdgeAI-Trust will deliver key innovations in technical areas for edge AI, including:
The EdgeAI-Trust project aims to leverage cutting-edge technologies to tackle challenges associated with ensuring the trustworthy and real-time orchestration of critical applications using decentralized edge AI within the Functional Safety Continuum. The project also seeks to validate the effectiveness of these technologies through real-world use cases.
The architecture will support the development of collaborative AI-based systems with orchestration, upgradeability, manageability, reliability, real time, safety, security, and energy efficiency. The architecture will support a continuum of heterogeneous AI-based algorithms and devices along standard APIs for interoperability and trusted exchange. This will range from sensor-actuation, large scale device-connected systems, edge processing units to the cloud. While the architecture is domain-independent, it will enable domain-specific instantiations for trustworthy collaborative AI systems.
Develop the next generation decentralized HW/SW edge AI technologies, with a particular focus on safety-critical and security related systems. The technologies will support fully collaborative AI by allowing heterogeneous devices to learn, adapt at the edge to cognitive reasoning tasks. Decentralisation of resources and processes among different entities/devices enable dynamic reconfiguration of processes in a resource constrained environment.
Considering the upcoming EU AI Act, here the focus is on explainability, reliability, safety, security and robustness of edge AI solutions. We will achieve that by focusing on (i) rigorous plausibility checks, (ii) monitoring of AI decisions and anomalies, (iii) explanation of situational awareness, and (iv) also include reliability and fault tolerance in a dynamic zero trust environment.
The goal is to provide methodology and tools that enable the optimization and validation of AI systems based on the EdgeAI-Trust architecture and solutions from the development platforms to the finished product. Automation supports developers by taking care of the constraints (like accuracy, latency, resource constraints and reliability) due to the limitations of the edge AI systems. By that, rapid improvement and deployment of Edge AI-based systems is provided.
Achieve large-scale, sustainable impact through the EdgeAI EDEM (EdgeAI Ecosystem Monitoring) platform while delivering a complete market analysis and an exploitation mechanism for financial sustainability and economic leadership in the global marketplace.
Dynamically monetize web-enabled processes through client-based action items. Authoritatively grow goal is oriented markets through ompletely generate technically sound content without robust users.
Founder & CEO
Cheif Expert
Head Manager
Founder & CEO
Pro Developer
Lead Developer
“The passage experienced a surge in popularity during the 1960s when Letraset used it on their dry-transfer sheets, and again during”
“Quickly formulate high yield web services before functional process improvements enable premier with e-business customer service.”
“Lorem ipsum, or lipsum as it is sometimes known, is dummy text used in laying out print, graphic or web designs the passage is attributed.”
The work plan has been designed according to the familiar V-cycle model. After working out detailed requirements and system designs in WPs 1 and 2, the main research and implementation work is conducted in WPs 3 and 4, before an integration phase in WP5 and the final validation and verification actions in WP6 are conducted. In parallel and over the whole duration of the project, WPs 7 and 8 cover dissemination, exploitation, standardization and project management respectively.
This work package aims to define the requirements for EdgeAI-Trust so that they drive the research in the project, as well as use case and validation activities. For every sub-system addressed in the project, there will be a task where the involved partners define the requirements and specification in close coordination. Industrial partners will steer the requirements engineering process based on their own experiences as suppliers but also on the insights from the requirements of their customers. Needs of the academic community will be brought in by the research partners. The task will gather top-down requirements (i.e., from the market, operator, cities pilots, and regulation, standardization and use-cases) and based on them, it will develop and specify technology-specific requirements for the implementation of the EdgeAI-Trust.
The overall objective of this work package is to establish a comprehensive, in-depth representation of the subsystems that will be integrated into the decentralized edgeAI ecosystem by conceptual design, modelling and simulation at different levels.
The overall objective of this work package is to develop essential semiconductor solutions on component level enabling EdgeAI-trust’s technologies that are implemented in the supply chains. This is based on the requirements and specifications that are defined in WP1 and in accordance with the system design development in WP2. The focus will be on Smart sensors and embedded control devices that will be integrated, e.g., by the EdgeAI-trust’s manufacturing, agriculture or the automated vehicles demonstrators. For that purpose, the objectives of this work package are directed in particular towards the following items:
The overall objective of this work package is to develop and implement collaborative edge AI and advanced embedded software modules for the supply chains. Thereby, these components cover different levels and application areas:
Development of the software components that are necessary to enable the collaboration and orchestration at all levels of the architecture (hand-over, load balancing of AI related tasks, model splitting, distributed/federated learning/training, etc.) to achieve more scalability, energy efficiency, robustness, low latency, as well as self-configuration, self-healing, resilience and data protection for the edgeAI solution, simplifying development, deployment, and operation process and life cycle management for connected adaptive mobility solutions.
The objectives of WP5 are to integrate the results of the EdgeAI-Trust project supply chains into the demonstrators. The SW and HW development outputs of WP3 and WP4 will be integrated and assessed in subsystem demonstrators to demonstrate the features in different vehicle functional domains. Different test set-ups including HW/SW will be implemented to demonstrate different ECAS vehicles functionalities needed to integrate the developed solutions, so they are ready to be demonstrated and tested on a test bench or a vehicle in WP6. The integration results in WP5 will be used to verify/validate the novel concepts, the requirements and technical specifications defined in WP1.
The objective is to integrate all the developed solutions, so they are ready to be demonstrated and tested on a test bench or a vehicle. System integration and partitioning is more and more crucial to assure highest energy efficiency, improved robustness, cost reduction and less dependency from the suppliers. Critical points have been addressed individually in the project in the previous work packages to provide innovative semiconductor-based solutions for increased reliability and availability, safety and security of intelligent systems, and improved functionalities. This integration phase will deal also with the subsystems and single components, both HW and SW, depending on the specific functions implemented in demonstrator.
It includes the final verification and validation (V&V) activities of each of the EdgeAI-Trust supply chains. The systems and sub-systems integrated and tested in WP5 will be verified against the requirements and validated in the use cases as defined in WP1. While testing on the unit/component level (the bottom-right side of the V-process) is part of the development done in WP 3 and 4, activities in WP6 focus on V&V of system/demonstrator level (top-right on the V).
Demonstrators as described in SC1-SC7 will be used to validate results in appropriate context depending on the target TRL (TRL5: demonstrate; TRL6: validate in “relevant” environment).
The contribution of SC8 in this WP are twofold: (i) SC8 will analyse each demonstrator of SC1-SC7 regarding the impact on trust and safety of the EdgeAI ecosystem as well as end-user acceptance; (ii), SC8 will validate the individual objectives defined in EdgeAI-Trust by assessing the improvements reached in SC1-SC7 beyond state-of-the-art.
The objectives of this WP are:
The objectives of this WP are the following: