¾«¶«Ó°ÊÓ

There is much to applaud in the government’s plan to turbocharge the integration of artificial intelligence into the UK economy, with ministers championing the technology’s potential to transform everything from healthcare to the planning system to fixing potholes. Construction should also be hugely excited about the AI revolution – not least because the technology could help boost the industry’s woeful levels of productivity, which have been such a long-standing problem and are linked to many of the other issues construction faces, such as a lack of investment in skills and innovation.

It is often said that construction is one of the few industries that allowed the industrial revolution to pass it by. Just look at housebuilding, where in many ways homes are still being built as they were 100 years ago, with contractors laying bricks in muddy fields.

There is an increasing risk now that construction will fail to reap the benefits of the technological revolution – and in particular AI. This is because construction remains dominated by a business model that sees a main contractor managing an ever-growing chain of smaller suppliers.

>> Also read: What to expect from AI in construction in 2025

>> Also read: Embrace AI technologies, Scape boss urges construction

When it comes to the application of AI, this is a problem, because AI is only as good as the data it is trained on. Data needs to be accurate, standardised and owned.

If the data – on everything from design standards, productivity, compliance and performance in use – is dispersed across a supply chain, it becomes very difficult to integrate AI effectively into a programme.

This means construction risks missing out on the enormous benefits that AI can bring. In partnership with Autodesk, Reds10 has been integrating a range of different AI applications into our construction programmes over the past few years – and we are seeing some remarkable results.

For example, we are now using AI to optimise cladding design for our projects, dramatically reducing the time required for this process. Design work that used to take eight to 10 weeks can now be completed in a matter of minutes.

This rapid advancement is possible because we manage our design in-house, giving us full control to capture, standardise and refine data continuously. By maintaining control over design and the process, we can systematically improve and enhance efficiency.

By integrating these technologies, designers can focus on creativity rather than worrying about buildability, clashes or manufacturability

If these aspects were subcontracted, achieving the same level of automation and optimisation would not be possible – at least not to the same level.

Another tool we have been exploring is Autodesk Forma, which allows for rapid exploration of multiple planning options and optimised layouts based on rule sets. While it currently focuses on mass modelling with low detail, the real potential for us lies in linking central manufacturing models with AI-powered generative design.

This could enable faster analysis, smarter designs and improved coordination across projects. By integrating these technologies, designers can focus on creativity rather than worrying about buildability, clashes or manufacturability.

Reds10 has also been increasing the amount of controlled and linked evidence stored in Autodesk Construction Cloud (ACC), utilising LOD 500 3D Models. For us, an LOD 500 model is not just for visualisation – it is machine-readable and can be directly interpreted by CNC machines, improving precision and enabling automation in fabrication and prefabrication.

It is not hard to imagine a future where centralised models automate the creation of detailed buildings based on predefined rules, streamlining the entire design-to-construction process

To enhance site documentation, we are also integrating AI-powered 360-degree cameras like OpenSpace, which automatically capture, map and analyse site conditions as teams walk through a project. These cameras provide continuous visual records, allowing AI to compare progress over time, detect discrepancies, and enhance quality assurance within ACC.

Reds10 can maximise the integration of AI because of the increasing levels of standardisation and industrialisation across our products, as we deliver sustainable high-quality buildings for the public sector, including homes for the armed forces, schools  and new buildings for the NHS and the prisons system. Just as consistent data is essential, high-quality, standardised models with structured information will be critical for fully harnessing generative design and AI in the future.

It is not hard to imagine a future where centralised models automate the creation of detailed buildings based on predefined rules, streamlining the entire design-to-construction process.

In many ways we feel like we are only scratching the surface of AI – the potential is so huge. But, for construction to fully benefit from that revolution, standardisation and industrialisation of construction through a more programmatic approach across the industry will be vital, ensuring further productivity and performance benefits.

Unless the construction sector presses ahead with the industrialisation agenda and reforms its broken business model, it risks missing out on the AI revolution – and on a better future, once again.