What have we learned?

CReDo is an ambitious cross-disciplinary project, and the team has derived many lessons from this process, resulting in a range of recommendations. It is important for future phases of CReDo to consider how the lessons learnt from the initial phase can be scaled up to bigger and more complex infrastructure systems. The methods used in the first phase should evolve to allow for greater scaling, repeatability and consistency across wider networks of assets.

Recommendations

  • Successful development of connected digital twins requires the involvement of multi-disciplinary teams. Connected digital twins cannot be developed by data scientists or domain specialists alone. Collaboration across disciplines is essential to achieving a connected digital twin that will fulfil its purpose.
  • A discovery phase is essential when the desired outcome and path of technical development is not clear. The vision needs to be set and a use case clearly defined during the discovery phase allowing for the involvement of representatives from all the relevant teams. As the Gemini Principles set out, a digital twin needs to have a clear purpose. A discovery phase needs to allow for technical freedom and exploration to identify what work is to be done and how it is to be done and what data is available before that project plan is translated into legal agreements. See the Legal Resources page for an example of license agreements to be set out at this stage
  • It is important to curate a data and model register from the outset to identify gaps and move to close them quickly or rescope the use case.
  • Ontologies help with scaling up and knowledge graphs can support ontologies.
  • When working to tight technical deadlines, it is important to keep a multi-disciplinary team motivated on the purpose of the digital twin. Public engagement events like webinars and public reporting help to maintain momentum and unite the team in collaborating on a common purpose – using data for the public good.

Legal

The CReDo Data Exploration Template is a copy of the licence put in place between CReDo Data Providers (Licensors). For CReDo, these were Anglian Water, BT and UK Power Networks, and for the Licensee, Science and Facilities Council (STFC) members: the Hartree Centre and DAFNI, who held the data securely on the DAFNI platform and managed access for approved parties to enable the technical development of CReDo.

THERE ARE MANY LICENCE OPTIONS TO CHOOSE FROM FOR SHARING DATA SECURELY AND SAFELY, WHY CHOOSE THIS OPTION?

This phase of the CReDo project was funded through an Innovate UK grant to the University of Cambridge’s Centre for Digital Built Britain and the National Digital Twin programme (NDTp). The NDTp set up a Consortium agreement to undertake the technical development of CReDo. Through the Consortium, the STFC provided secure data storage and managed access to the data for permitted parties, and to the work that would require access to the data. Therefore, liability for unauthorised access or use of the data lays with STFC.

The Consortium and Participation agreements signed by the rest of the CReDo technical delivery team, the Permitted Third Parties in the licence, included contractually binding provisions that reflected those in the licence, sharing the liability across all parties.

To keep the legal overheads on CReDo as light as possible, one licence was issued jointly to STFC by the three Data Providers.

COULD IT HAVE BEEN DONE DIFFERENTLY?

Initially, we approached the Data Providers to arrange individual licences with each party that would be accessing the data, however this would have brought with it the need to administrate a large number of contracts and the overhead for the project, in which the Data Providers were taking place voluntarily, was deemed too high.

WHAT ABOUT DATA TRUSTS?

Data Trusts are still in their infancy and require legal expertise and funding to establish. In the context of this phase of CReDo, a data trust was not considered, as delivery of the Climate Resilience Digital Twin was restricted by the available budget for legal investigation and timescales for delivering the required outputs of the funding arrangement. Future projects adopting the CReDo approach may consider pursuing this route.

WHAT IS A DATA TRUST?

The Open Data Institute (ODI) defines a data trust as:

a legal structure that provides independent stewardship of data”,

and has written a report called Data trusts: lessons from three pilots that describes a set of pilot data trusts carried out by the ODI and the framework it has developed for setting up a data trust. A sister paper, written by BPE Solicitors, Pinsent Masons and Professor Chris Reed, Queen Mary University of London, Data trusts: legal and governance consideration, states:

One important conclusion from our research is that each data trust will need its own, individually designed, legal structure. It is not possible to recommend any single form of legal structure or even to produce a set of templates from which data trusts could choose.

For future work on the CReDo project, it may be possible to create a data trust that supports the development of extended data sharing, but it will need time, legal resources, funding and the establishment of a board of trustees or data stewards, separate from the data providers and users.