In construction and infrastructure, digital transformation is much deeper than just counting the number of platforms deployed, models created or dashboards launched. If that is your metric of success, you could be creating more data without gaining more knowledge.
When trained or prompted with unreliable information, an AI system will give unreliable assistance more efficiently.
For example, duplicated files, inconsistent metadata, unclear approvals and decisions made by email can still exist within a project in the cloud. It may use building information modelling yet provide information that an asset operator cannot search, trust or maintain. It’s typically not another application that’s missing. Information discipline is essential.
In this article, I will look at what is holding back the sector from truly realising the benefits of digital, highlighting key principles I follow that have worked, as well as setting out my five tests for digital maturity for businesses.
Key principles for a successful information strategy
- Trust and control are key – Technology leaders need to achieve trust at scale: to be confident that the information they have in front of them is up to date, authorised, has sufficient detail for the purpose and is traceable to a responsible source. That may seem simple, but it requires several small controls to be in place: agreed naming and classification, meaningful metadata, defined information states, controlled access, defined review roles and responsibilities and an auditable trail of changes.
- Repeatable governance not just new folders – The Common Data Environment (CDE) is often referred to as the “single source of truth” but it can breed false confidence. Not everything users can access is authoritative. A CDE is more than just creating folders. Permission should be based on roles. Metadata should be searchable and support assurance. Routes should be proportionate to the risk and complexity of the project. Naming rules should be ‘practical’ for consistent use. Getting these fundamentals in place through repeatable governance removes the need for avoidable setup decisions and variation, allowing teams to start with an agreed operating model.
- Getting automation right – Automation is great for repetitive tasks in information management: drawing information validation, updating, routing submissions, permissions, tracking deliverables and creating progress reports. It can reduce administrative friction and provide more consistent evidence. On one project, I used a Python-based solution to update drawing frames from controlled attributes, saving 80% of manual effort. At WSP, I delivered a low-code Power Platform solution for submittal tracking, reducing response cycle times by 65%. Those gains were not achieved by automating all of the potentially automatable steps. They started out by determining what data was needed, who was accountable for it and the exception route.
- Dashboards – Dashboards can be helpful in showcasing design progress, deliverable status, reviewing performance and identifying emerging risks. However, with the definitions being unstable and source data unreliable, a polished visualisation can’t hide that. Organisations should have an understanding of the meaning of each of the measures, the sources of the data, the responsibility for data quality and what actions will be taken before building a dashboard. Otherwise, the dashboard is an appealing summary of inconsistent information.
Five tests for genuine digital maturity
In my experience, there are five practical tests that technology leaders can use to determine the robustness of information management:
- Is there a way for the users to view the current authorised information without inquiring with the person who created it? Otherwise, the organisation has access only to the data, not reliable information.
- Do information needs relate to a decision or life cycle need? Data without a purpose drives up costs and compromises data quality.
- Are the following visible through the workflows: status, ownership and exceptions? When it’s possible for responsibility to fall between systems or teams, a process is not controlled.
- Does automation contribute to less friction and maintain accountability? Efficiency shouldn’t be at the expense of interpretation and oversight.
- Will the information be of continuing value after the project team has gone? Digital value is more than just delivery; it’s operation, maintenance and future change.
Information discipline for digital infrastructure
The growing use of AI, digital twins and interconnected platforms will add to the capabilities of organisations. They can also exacerbate the effects of poor governance. When trained or prompted with unreliable information, an AI system will give unreliable assistance more efficiently. A digital twin without information from controlled assets can be impressive, but not very robust.
It’s not always organisations with the largest software estates that will reap the rewards from new technology. These will be the ones that can precisely articulate information needs, have trust and confidence in systems and use information to make timely decisions.
Information management is not an administrative abstraction of digital delivery. It is the infrastructure that enables digital delivery to scale. When information is structured, governed and usable, technology is more than just a collection of tools; it’s a reliable method of working.
Olufemi I. Akinwumi is a senior information manager at WSP in the UK with over 8 years of experience across architecture, infrastructure, energy, digital engineering, and information management.
