Data Ecosystems: Sovereign Data Exchange among Organizations

Summary

The design of data ecosystems, infrastructures for the secure and reliable data exchange among organizations, is considered as one of the key technological enablers for digitalization and the digital economy of the future. Several applied research initiatives and industry consortia provide substantive evidence of this trend e.g., the Industrial Internet Consortium (IIC)formed in the USA, the Industrial Data Space (IDS) founded in Germany and the associated consortium International Data Space Association (IDSA)2. Most of these initiatives aim to provide a reference architecture for dealing with (i) governance aspects related to the definition of policies and conditions able to norm the participation to the data ecosystem, (ii) security aspects related to the definition of policies and infrastructures for guaranteeing a trusted and secure exchange of data, (iii) data and service management aspects related to representation models and exchange formats and protocols, and (iv) software design principles related to the realization of the architectural components and their interaction.

All these aspects have been discussed in the seminar and the main findings are described in this report. In addition, a central new aspect of data ecosystems that we considered in the seminar lies in the view of data as having an economic value next to its intrinsic value to support operational and decisional core business activities. This means that in the data ecosystem, data is typically considered both a business asset and a business commodity which may be priced and sold in some form (e.g., data provisioning service or raw data) according to contracts.

As testified by the amount and variety of problems described above, the creation of such ecosystems poses many challenges cutting across a wide range of technological and scientific specializations. For this reason, the seminar involved researchers from different communities. Interdisciplinary discussions gave the possibility to analyze different perspectives and to achieve valuable outcomes presented in this report, such as a wide set of research challenges and the definition of interesting use cases for the further development of data ecosystems. Details about the activities carried out during the seminar are provided in the following.

Overview of the activities

The seminar took place from Monday September 23 until Friday September 27. The seminar program encompassed four invited talks (keynotes and tutorials) on the first day (Sep. 23rd), by Gerald Spindler (law and ethics), Frank Piller (ecosystems and business models), Maurizio Lenzerini (data integration), and Boris Otto (International Data Space). After discussions related to the talks and tutorials, the remaining afternoon was spent structuring (through joint discussion) the coming days of the seminar and group structure. As a result, group structure was based on a thematic structure encompassing three groups, one for each of the topic areas Business, Data, and Systems. Tuesday Sept. 24 began with a breakout into groups and election of scribes in each of the three groups (Business, Data, and Systems), and the remainder of the day was taken up by parallel group sessions in the three groups. Wednesday Sept. 25 began with a joint session where each of the groups presented their work, which was then discussed jointly. The afternoon (until the excursion) was taken up by joint discussion on report structure. The morning of Thursday Sept. 26 encompassed joint discussion on a proposed joint manifesto as well as group discussions on application domains and application scenarios (topic areas were Health, SmartCities, Industry 4.0). The afternoon was taken up by continued group discussions and ended with group presentations and joint discussion on application domains and application scenarios. There was also further discussion on report structure at the end of the day. The manifesto was subject to very lively discussion in the evening, after dinner. Friday Sept. 27, the last day of the seminar, was devoted to wrap-up (conclusions, summary, and report process) followed by joint discussion on relations between Systems, Data and Business views on the overall topic of the seminar.

The outcome of the seminar, which is documented in the remainder of this report, encompasses summaries of the group discussions and the joint manifesto.