Kier, alongside Cambridge University and nPlan, a company which uses machine learning to analyse construction schedules to determine risk and recommend improvements, has recently been awarded an £846,000 grant to deliver a research programme into the use of artificial intelligence (AI) to better predict, plan and schedule construction projects. So, will the use of AI fundamentally change how we deliver projects? Chloe Anfield, head of innovation at Kier and the company’s lead on the programme, explains.
“As an industry, we have a reputation, rightly or wrongly, for projects overrunning. There are many, many factors that can affect project delivery, such as unexpected issues with the site and local factors, changes from the client, logistics and access to busy city sites. Our research programme uses real data from across the industry to train an AI machine on likely project delivery timelines, based on historic delivery times from across the industry.
This will give us a more realistic idea of the risks to delivery times and when the project is likely to be delivered. This could help with client negotiations, which will be based on industry-led fact, adding credibility to our position.
As well as understanding how good our project planning estimates are, we can make data- and fact-based decisions about the commercial viability of projects at the bidding stage. Once we are in the delivery stage, we will become more efficient, leading to reduced project costs. I am personally very excited to see what AI does with our project schedules; will it find something ground-breaking for construction?
Over time, we are looking to embed the solution with change impact, which will show the real cost of changing project delivery schedules. We will then be able to reduce contingency-based cost and build greater confidence in the industry’s ability to deliver on time.
Ultimately, we’re hoping the programme will help reduce costs, increase efficiency, improve productivity, performance and collaboration, and improve the image of construction, all of which will lead to economic growth and increased profit.
Kier is in a great position to support the project’s development, ensuring it is applicable to the real world. We’ve just completed the first quarter of the two-year project, and – as part of our information gathering to help with machine learning – we’ve completed interviews with Kier planners and project managers to understand the real-world challenges of project delivery.
The research team recently headed to Hinkley Point C to learn more about how planning sessions work in practice and the complexities that can affect decision making. The team came away with some interesting findings, which we will be sharing in the coming months.
The project supports the Industrial Strategy Challenge Fund, which strengthens UK science and business innovation and takes on the biggest challenges that society and industry face today. I hope it is the first of many collaborative projects we can support to advance our industry.”