The second in my series of project data analytics blogs looks at data trusts… last week, I outlined the opportunities that could be realised if we begin to leverage the exhaust plume of data that emerges from projects. Creating a data trust, that could securely capture that data across the industry, would provide a fundamental part of this jigsaw. If we don’t create appropriate levels of quality data, we will struggle to realise the benefits that advanced data analytics could potentially offer.
So, what is a data trust?
A data trust is a legal structure that allows organisations to securely pool their data to address challenges that all or some parties may have.
By pooling data we can assess our collective ability to answer the questions that we demand of the data. There will undoubtedly be a shortfall. This will grow as we understand the potential of advanced analytics and ask even more demanding questions; as we begin to see life through an AI lens. But by working together we can improve how we harvest data (using automated methods where possible), align the scope of the data with the questions we are answering and also ensure that quality and consistency is appropriate. Collaboration is key to success and we can build on work being undertaken elsewhere.
The Open Data Institute has been leading some of the thinking on data trusts, but their applications are mainly aimed at open data, i.e. providing public access to data. In contrast, the Grampian Data Safe Haven - a joint facility between NHS Grampian and the University of Aberdeen - provides highly secure anonymised access to medical data for research on the factors that influence health and disease over the course of life.
Our challenges are slightly different, an open data model will probably lead to data being redacted, anonymised or withheld, likely impacting the achievement of the vision, while being overly secure and anonymising all the data may be impractical, unnecessary and overall complex. The opportunity lies in building something that suits our needs.
Is there ‘trust’ in the data trust?
One of the first things we must grapple with is how we encourage data providers to share their data. Some data is commercially and reputationally sensitive so we need to put appropriate controls in place to ensure that it isn’t misused. If there is a probable risk of the data being used against the interests of those providing it, then they simply won’t release it.
Trustworthiness is central to success. It’s something we need to earn. This prevents the data trust from being open data, but that doesn’t mean it’s a closed shop.
Essentially, data providers will input into a trust; their interests will be represented by trustees, which may also be nominees of their organisation. The data steward will then integrate the data into the overall model, assessing the risks associated with releasing the data and how some of these risks can be mitigated. Third parties can then apply to access the data in the trust for defined purposes. Academics are likely to gain access to raw data, with conditions. Conversely, competitors will have restricted access to anonymised data. We will need to adjust access based on a number of parameters. Ultimately, it will be a decision for the trustees on who can access their data and for what purpose. But if we overly constrain access then we don’t deliver the MER goals and delivery productivity will suffer. We need to find the balance.
I’ve had a couple of people ask me whether organisations will share their most sensitive data. I have my own views on it, but this is irrelevant. What’s important is that we need to test it and flesh out the viability of an energy industry data trust and the benefits that will derive from it... starting with data that we are all comfortable sharing and then expanding as we gain confidence.
Such a trust has the potential to transform how we deliver projects. It also enables operators and contractors to share good practice, leverage their collective experience, improve delivery confidence and reduce cost. History has shown that we struggle with the ‘lessons learned’ process. By simply changing our language to leveraging experience, we open up a whole host of data-enabled opportunities. A data trust is core to the realisation of this vision.
How close is industry to realising this vision?
The OGTC is currently undertaking a project looking at the legal and governance issues that need to be addressed, and have developed a working framework. Our ground-breaking work has also enabled similar initiatives in other sectors. It’s worth checking out the datatrust.construction website - it’s an initiative that is gaining momentum, but there is an opportunity for the energy industry to take a lead.
This work has also value to other areas where the industry could benefit from a legal and governance framework for pooling data for the development of AI solutions. We are at the start of an exciting journey.
How will you and your organisation be part of it?
Let’s transform how projects are delivered… by working together.
We need project professionals to review and provide input before we move on to the next phase and start capturing data. If you are interested in getting involved in any aspect of the project, please get in touch by contacting firstname.lastname@example.org