Collaboration, new skills, removing barriers to sharing data and rapid experimentation all form part of the data analytics and AI plan, announced by the government’s Project Delivery Function.
The government’s Central Digital & Data Office (CDDO), the Infrastructure and Projects Authority (IPA), the Association for Project Management (APM) and the Major Projects Association (MPA) are involved in delivering the plan.
In its announcement, the government states: “Combined with the uncertainty surrounding the outcome of rapid technological change, this means that a top-down, one-size-fits-all approach to exploiting data is neither possible, nor desirable.
“Instead, government and its partners will collaborate to create the right conditions for innovation to thrive and ensure that success is shared at scale across the projects ecosystem.”
The first tranche of actions are grouped around five themes, as follows.
1. Data skills and capability at scale
The Government Project Delivery Capability Framework, and the associated government accreditation scheme, will be updated to clarify the skills and roles needed for the future. Working with the APM, the government will set out interventions to grow and deploy these skills.
The government notes that there remains uncertainty about the impact of data analytics and AI on existing project delivery roles. Some may be significantly impacted, it said, with role redefinitions required to include data analytics responsibilities and competencies, while entirely new roles are also likely to emerge.
There will also be implications for project delivery models. Projects will need to consider the extent to which they draw upon third-party data specialists against growing their own talent, either centrally in the portfolio or embedded within the project team. As a minimum, having enough data skills in-house will be critical to being a ‘smart’ procurer of services and tools from the market.
2. Better data and availability
The government and industry will work together to develop a common set of standards and taxonomy for project data, building a foundation of FAIR (findability, accessibility, interoperability, and reusability) data. Through this, expectations will be set on rights of ownership and access to data.
3. Evidence-based decision-making
The government will “strategically and systematically” build the infrastructure across government project data and tools to enable greater insight, benchmarks and the ability to predict performance.
The IPA is in the process of moving its government major project data collection to a purpose-built Cabinet Office platform. This will enable data collection tailored to requirements.
Meanwhile, the IPA's new, cloud-based Benchmarking Data Service is will enable departments to learn from historical data to better inform decisions on current and future projects. It will allow benchmarking at various levels – from individual projects to different types of project and the portfolio as a whole.
4. Experimenting together
A series of pilots to innovate and experiment will be run, with “the best bets amplified and scaled at pace”.
5. Data partnerships
The government will continue to work with professional bodies, academia and industry, focusing on the shared objectives.
Delivery against these commitments will be a collective effort led by the IPA, as the centre of expertise for project delivery in government. It will work with the Projects Council (comprising the chief project delivery officers of central government delivery departments and chaired by the IPA chief executive) and heads of profession in departments, the CDDO, professional bodies, academia and industry.
Defining data and analytics
The plan defines project data as “any data used for selecting a project and for defining, monitoring and tracking its performance”. This could include project registers, schedules, plans, budgets and forecasts, meeting minutes, reports, and assurance findings. This definition will be refined to create a more detailed definition of project data.
Project data analytics is defined as using project data to:
- automate routine project tasks;
- predict future project performance; and
- help make better project decisions.
What’s the opportunity?
“Better outcomes,” the government says. Given that this plan follows hot on the heels of the publication of the lessons learned on Crossrail, it is timely for the government to state: “Systematic aggregation, sharing and learning of lessons across the portfolio could improve outcome-focused decision-making and approvals. It will help us set more realistic goals and temper optimism bias. It also presents the opportunity for better option selection, and indeed better project selection, allowing us to focus on those interventions, which give us the biggest return on our investment.”
The risks
The plan acknowledges the risks. “Perhaps the most immediate risk of the application of AI in project delivery is that insight from data is not fully understood, based on poor-quality data or incorrectly applied models,” the government says.
It notes that the approach must also remain alert to the longer-term risks posed by AI, but “in the short term, given our anticipated use cases and current level of maturity, the adaptiveness and autonomy of AI pose fewer immediate risks for project delivery, but will be closely monitored as the technology develops and new use cases emerge”.
Furthermore, the government says that doing nothing is not an option, but to try to anticipate how the technology will develop and iterate would be counter-productive.
Don’t miss out on BIM and digital construction news: sign up to receive the BIMplus newsletter.
