By Andy Cunningham, Senior Regional Director for Australia and New Zealand at Autodesk.

The construction of the Sydney Opera House signalled the birth of a global architectural icon, but it also demonstrated the potential for disaster, according to author Andy Cunningham.

It took 10 years longer to complete than expected, cost 10 times the original budget, and was riddled with design challenges and poor collaboration – as well as significant bureaucracy.

Now imagine the architects, engineers and builders working on these projects had their expertise supplemented with AI.

It would facilitate collaboration and streamline workflows by summarising vast amounts of project data from multiple models, empowering all parties to work together more effectively using previously untapped insights.

Project management and planning would have been dramatically enhanced with predictive timelines based on historical data; real-time information analysis would identify risks, safety hazards and quality issues before they become critical, and cost estimation would be automatically optimised.

Meanwhile, designs would have been generated in granular detail with consideration for budgets, environmental factors and space utilisation, and finally, predefined standards and regulations would be baked into the entire process.

The construction sector is often criticised for being a laggard in digitalisation. While we can’t shy away from the truth of the past, data shows attitudes are changing, and that the sector is drawn to the role AI stands to play.

A recent report revealed two thirds (67%) of leaders and experts from design and make companies – including Architecture, Engineering, Construction and Operations (AECO) – believe their organisations are approaching or have already achieved their goal of incorporating AI into their operations.

But it’s important to note these figures represent AI in its many forms and use cases. While ChatGPT popularised the role of AI in everyday society, it takes a plethora of different forms in the context of AECO, dictated by the requirements of each individual sector, as well as the needs of cross-sector collaboration.

That could include machine learning, deep learning, generative AI, or generative design.

Similarly, AECO industries sit at different stages of a broad spectrum of maturity.

Many architects, for example, are already accustomed to powerful AI image generators, which merge the craftsmanship of their highly skilled workers with digital renders that leverage existing images and text commands.

On the other hand, many builders are earlier in their AI exploration, currently creating foundations for ChatGPT-esque capabilities that rapidly resolve enquiries that would otherwise have taken various stakeholders days or weeks to coordinate by manually sifting through pools of documents.

What’s almost universal, however, is that the increased accumulation of unused project data highlights the potential of leveraging AI to glean insights that will drive innovation, simplify collaboration, accelerate sustainability initiatives, plug talent shortages, streamline costs, and save time.

The aforementioned report notes 79% of organisations trust AI, and that 67% agree it will be essential across the board in two to three years. In my discussions within the industry, the ability to keep up with the flood of data and general demand – more buildings, services, products, content and so on – is precisely why AI is not only seeing significant interest, but also investment.

But the benefits of AI won’t be instant and AECO organisations must first focus on establishing an operational framework to ensure it can deliver outcomes. It takes a pragmatic and structured approach.

That approach begins with establishing the right data infrastructure. Robust data management is critical as AI is a case of ‘garbage in, garbage out’ – the quality of the AI relies entirely on the calibre of information from which it draws.

The data piece must be backed by iterative implementation. As with most digitalisation strategies, going ‘all in’ is fraught with risk, and can cause disruption – particularly if its introduction doesn’t go exactly as planned.

Beginning with small, manageable projects creates capacity to test, learn and amend AI implementations based on real-world feedback and outcomes.

An adjacent and equally crucial consideration is where that data is kept. Digital projects live in the cloud, and while data centres around the world are built with security and performance front of mind, opting for local and on-shore options increases control over sensitive project information, and reduces latency for those who work with the data.

These two concepts need to be met with the most important part of any technology: the people using it. While some people are concerned about AI and its impact on their personal and professional lives, it will inevitably become a ubiquitous part of life – but it can only deliver intended outcomes if it’s geared to outcomes for stakeholders.

Employers therefore have responsibility to deliver ongoing education and training for their teams to ensure AI can supplement day-to-day tasks, and ultimately drive productivity by taking on low-value administrative work – it’s a collaborative relationship between people and systems.

The optimism towards AI is promising, as is the appetite to overcome the challenges that have burdened AECO industries to date.

However, for AI to deliver improvements from the design to the delivery of construction projects, organisations must be cognizant of the strategic scaffolding required to bring it to life.

Image: Wkikpedia Commons

Clémence Carayol

AI-in-construction-demands-its-own-scaffolding-to-1732061000.png

Let’s connect!