by Sascha Chandler, Integrated Infrastructure Partner – Infrastructure Risk and Controls, PwC and Alastair Pearson, Integrated Infrastructure Partner – Infrastructure Data and Analytics, PwC.
Intelligent risk identification and modelling will be key to successfully delivering infrastructure projects in these times of great demand and significant headwinds.
COVID shutdowns, geopolitical uncertainty and extreme weather events have slowed productivity and brought flow-on effects such as escalating input costs, shortages of skilled labour and, ultimately, higher levels of delivery risk for infrastructure projects.
Numerous large, complex, publicly funded projects have already been delayed and run over budget, and there will no doubt be many more projects yet to transition to delivery that will experience significant cost and time pressures.
Steadying the ship and setting a course for the successful delivery of these projects is crucial given that infrastructure projects stimulate the economy and employment opportunities and bring many other important benefits to our communities.
With increasing inflation and constraints on budgets and resources, managing the critical risks of our large infrastructure projects will be key to managing costs and bringing projects in as close to budget as possible, or understanding what tradeoffs may be needed.
To identify, assess and mitigate risks, the infrastructure sector will need a disciplined and well-planned mix of:
- The use of advanced analytics to help provide insight from the increasingly large, complex data sets produced by infrastructure projects and assets, to help make better informed decisions
- Robust governance, controls and assurance applied to planning, cost estimation, risk prediction, procurement and contracting
- Industry collaboration enabling risk, cost and performance data to be shared
Intelligent, data-driven insights
The field of advanced data analytics has surged ahead in the last decade and many industry sectors are recognising the benefits. Data elements related to risk, cost (e.g. timesheets, claims, material inputs, etc.), interfaces and delivery performance are readily available across any well-controlled infrastructure project. Technologies such as optical character recognition, high-powered computing, internet-of-things devices and high-speed data networks enable this data to be ingested, normalised and prepared for integration with sophisticated risk modelling and ‘what if’ scenario analysis.
This data and the insights it drives are also valuable inputs to ‘digital twin’ models, which are now being employed frequently when modelling multi-faceted infrastructure projects. Digital twins help assess the full value capture of major public investments and commissioned assets so that they are fit for purpose and achieve targeted community outcomes.
Increased computing power at lower cost now gives us the ability to build much larger and complex digital twins that can analyse data and ‘what if’ scenarios at a level of detail never seen before. This will continue to evolve in the coming decade with the advent of further advances in quantum computing, which will revolutionise digital twins.
Broader and deeper data sets allow more to be done and more to be revealed. Patterns and themes can be identified in the data by exploring the correlations between the data sets and the overlay of these correlations to time horizons. These patterns and themes can then be used in Monte Carlo simulations to help predict how various aspects of risk will impact on individual projects, portfolios of projects or the broader industry. As the data set grows in depth, breadth and history, the power of the analytics increases, as does its value for predicting risk.
Applying the advanced analytical techniques of artificial intelligence and machine learning to these broader and deeper data sets provides us with greater potential to move into more sophisticated predictions of outcomes based on historical patterns and contextual information. With enough data, predictions can be converted into prescriptive actions, auto-suggesting the best course of potential actions following a disruption or unplanned event in an infrastructure project’s development.
If applied across the industry, advanced data analytics can provide the information necessary to determine the best balance of risk share across the industry, and could inform infrastructure strategy and government policy. A shared and consistent understanding of risk between the public and private sector would surely lead to better delivery collaboration and ultimately reduced delivery risk.
Robust governance and controls
The governance, controls and assurance frameworks currently employed on major infrastructure projects have been developed to provide oversight of public investments and to increase the confidence of stakeholders. These frameworks can benefit greatly from project-specific, data-informed analytics.
Assurance practices can transition from a backward-looking, one-size-fits-all approach to a forward-looking, risk-informed practice. Trained predictive models informed by a rich historical performance and risk data set can help to identify early whether delivery may blow out, enabling proactive and targeted intervention to rapidly correct the course of the project.
When forward-looking insights are available at an all-of-industry or portfolio level, they can be used for comparative analysis between projects and sectors – thereby providing momentum to insights-driven continuous improvement activities such as targeted staff training, policy and process refinement, project-to-project knowledge sharing, procurement processes, and health and safety practices. These insights are also valuable inputs to community consultation and progress communication.
Industry collaboration
Gathering data from the many parties in the infrastructure ecosystem is perhaps the biggest hurdle to establishing deep and broad historical and current risk and performance data sets. Yet concerns related to ‘at the source’ data formatting and the commercial sensitivity of data can and should be overcome.
It’s not necessary to transition all projects onto consistent project management platforms – that would be too great a challenge. Instead, the heavy lifting of staging and normalising the data can be performed centrally by a portfolio management office or, in the case of industry-wide data collection, by the state or federal infrastructure oversight bodies.
The following factors will be critical for industry collaboration:
- Project and industry benefits will need to be considered carefully so that they can be communicated succinctly. Recognition of mutual benefits will help drive collaborative behaviours
- Expectations relating to data sharing should be established in supply contracts.
- The intended use of the data and its relevance to driving insights and avoiding risks must be understood and communicated and be central to platform design
- The value of the insights must be understood by industry participants who can collectively assist in informing design and expected outcomes
- Collaboration will be needed among big data and risk modelling specialists to ensure well-managed showcasing of design, construct and build and implementation of functionality within trusted and secure platforms
The enabling technologies for intelligent, data-informed risk identification and modelling are available, the market has appetite, and the benefits to the industry and ultimately the community are many. With strong and persistent sponsorship to coordinate all the moving parts and engage the industry, the vision of better infrastructure risk management and efficient, on-budget delivery of the infrastructure pipeline can become reality.
