The project has been progressing according to the work plan during a period of 36 months (from M1 to M36; April 2021 - May 2024).
A brief overview of the work performed per WP is:
WP1: This WP has been the project coordination since the KoM through the 36 months duration, providing support for all the administrative activities (i.e. quality assurance, configuration, data management and support for procurement and reporting), close management (i.e. monthly Project Board Meetings and Advisory Board coordination). Specific technical meetings have been promoted when necessary.
Organization and preparation of project reviews (RV1 in Feb’22 , RV2 in Jan’23, RV3 in Oct’23 and RV4 in April’24) in the context of communication with the Commision. The work done in this WP has permitted to fulfil the committed schedule and the objectives according to the DoA.
WP2: The WP has provided use cases description, requirements and real data provided by partners. This WP represents the perspective of a satellite or telecom operator taking advantage of the ATRIA functionalities. The WP2 activities have concluded with the onsite validation of the ATRIA AI-PCS tool (Milestones MS4-OSAT1 and MS9-OSAT2). Last milestone, in Feb 2024, has been an important achievement, since meant the validation in an operational environment and with the presence of the early adopters who provided very valuable feedback and important information for the future evolution of the prototype.
WP3: This WP consisted of the AI-PCS prototype, including the definition of the tool interfaces, the development of the web-based tool and the integration of the AI/ML models developed by WP4 to cover the four use cases selected. The validation of the AI-PCS tool also supposed the integration of the Flexible Payload Emulator (FPE) developed by WP5. The validation of the AI-PCS tool in lab/factory environment (Milestone MS4-FAT1 and MS8-FAT2) was performed at GMVs facilities as per defined validation and verification plan, completing all the WP3 objectives.
WP4: This WP consisted of the development of the AI/ML solutions to fulfil the requirements and use cases proposed by WP2, iteratively producing different versions versions of the algorithms for the four cases considered. It has paid attention to the quantification of the validity of the results carrying out AI model validation tests prior to the integration of the models in the AI-PCS tool.
WP5. This WP has been focused on the definition, design and implementation of a FPE with beamforming capabilities to validate the correct behaviour of the AI-PCS in a laboratory environment. The FPE was integrated in the AI-PCS tool validation setup and this WP gave the necessary support to WP3 to complete this activity.
WP6. This WP has collected the market and business opportunities for the project with especial remark to the early adopter’s feedback collection and the preparation of the Business and Exploitation Plans preparation for the project outcomes. This WP has also managed issues related to the industrial property rights for the mentioned produced results.
WP7. This WP has performed the dissemination, communication, standardisation and the technology transfer of the project results. The WP has worked in several aspects and initiatives, as planned, such as scientific and academical events (e.g. organization of industry panel of the SDPNGS 2023 workshop), industrial events (e.g. Satellite), social media (Linkedin, Twitter) and the project web page update (
https://www.atria-h2020.eu/(odnośnik otworzy się w nowym oknie)) publishing all the results and the deliverables of the project (public ones). It also included the production of a MOOC (massive open online course) as mean to transport the scientific work into industry.
The variety of exploitable products obtained from the project have been: AI/ML and training models prototypes of KONNECT (Product), AI/ML models for phase 2 (Product), AI/ML and training models prototypes (phases 1 and 2) (Product), FPE (Research), Synthetic dataset from payload emulator (Research), Telecom Flexible Payload MOOC (Research) and AI-PCS tool (Product).
Depending on the concrete asset, ATRIA proposed different ways of exploitation, such as use for research activities or commercialization as a product.
