10. Conclusions
This document details the elements of a data sharing framework in order to facilitate re-use of the many CCAM/FOT/NDS datasets hosted at different locations globally. This framework can also facilitate data sharing within new projects, as the content of the framework is general and could be used whenever data sharing is performed. The framework introduces the new approach of federated data sharing, applied to a CCAM domain.
The CCAM Data Sharing Framework then consists of the following seven items: project documents (such as the consortium agreement and the consent form), description of data and metadata, data protection, training on data protection, support and research services, financial models for post-project funding and data governance as described in application procedures. All these components need to be considered for efficient data sharing framework.
The report constitutes the essence of the discussions held during the European support action projects FOT-Net 2, FOT-Net Data, CARTRE, ARCADE and FAME time frames. Through the discussions, it has become obvious that the recommendations apply to a wide variety of cases and research areas (applicable but not limited to automotive), including different national contexts. At the end, though, it is always up to the partners of the specific project, national or international, to select the appropriate data-sharing strategy and decide what parts of the data sharing framework are applicable to their project.
The CCAM Data Sharing Framework needs to be discussed and applied by different stakeholders with good knowledge of their national requirements, who collect experiences and make the framework applicable to as many countries as possible. The framework also needs to be updated as new technology and methods provide new possibilities, especially regarding anonymization and feature extraction. Reliable tools for automated feature extraction are key to be able to provide large quantities of essential features from video.
If the suggestions and requirements presented here are taken into consideration, future projects will be well prepared for data sharing during and after the project. Sharing data can be a win-win opportunity for both those who share and those who analyse data. The Data Sharing Framework has been applied by many European projects and organizations since its release in 2016. The additions of this version will help to set the foundation for a data space to support future projects within connected, cooperative and automated driving.