How AI can transform construction

Productivity is one of the central challenges for innovation in construction today. While our industry has fundamentally transformed operating paradigms and impact through remarkable progress in safety protocols, environmental stewardship, and social responsibility, improving productivity has remained persistently challenging. As the Australian Constructors Association's (ACA) 'Disrupt or Die' report illuminates, this productivity plateau has had to be overcompensated through increased labour and capital inputs, with consequences such as increased construction costs and, in some cases, project delays. The ability to build our environment is a core capability of our society, directly impacting the quality of our infrastructure and ways of living, so improving productivity through innovation is of foundational importance.

The trajectory to improve productivity appears promising. As we navigate 2024, the construction industry stands at an unprecedented junction of opportunity for achieving step-change productivity improvements. Our sector has demonstrated strong interest in exploring these possibilities, and it’s at the cusp of understanding how emerging technologies, such as Artificial Intelligence (AI), can help us transform the way we build. With AI, we can address some of the core issues of our time and ignite a transformational change in the construction sector.

Our industry's capacity for technological adoption provides further grounds for optimism. Counter to prevailing perceptions, the construction sector has historically demonstrated remarkable innovative prowess, leading to broad transformations on how the industry operates. Consider the evolutionary trajectory of tools such as power equipment and hydraulic machinery, advances in building systems like steel-framing, developments of new materials like high performance concrete; each of these has led to a fundamental transformation of how the industry works and what it can deliver. More recent technological advances, including drone technology, Building Information Modelling (BIM), and prefabrication methodologies, continue their gradual but steady integration across various industry sectors. In this context, AI represents the next chapter in our industry's long history of transformative innovation.

The imminent potential of AI

AI development has recently accelerated, and several AI-based tools can now demonstrate productivity improvements in all parts of the modern construction process. AI, while most broadly known today through technologies such as Large Language Models (LLMs) and generative AI, represents a broad suite of statistical and computational methods through which algorithms can understand something about the world and the humans living in it.

Productivity improvement from AI for the construction industry is expected to come from two directions: from its ability to shoulder the burden of the mundane, routine and ordinary, and the ability to see through complexity and shed light on implications and insights that were previously hidden to us. What do each of these abilities look like in practice?

Let’s start with the first. Modern construction is characterised by an extensive and complex process managed across numerous interfaces, and workers at all levels end up spending significant amounts of their time managing what can be called ‘foundational’ tasks. These are tasks that are important and must be executed well but are not the things that will differentiate a project. Routine correspondence, repetitive administration tasks, cross-checking data, filling forms; much of the time of the modern construction worker, in all roles from project engineer to bid leader, is filled with such tasks. Each minute spent on a foundational task is a minute that cannot be spent on a value adding task – finding creative solutions, building connections and finding opportunities to drive improvement.

Early research suggests that AI can substantively shift the balance between value-adding and foundational tasks in construction, with reductions in time of 65% spent on foundational tasks being demonstrated by some companies. By automating routine administrative work, AI liberates human potential, allowing skilled professionals to focus on creative problem-solving and innovation. For example, Large Language Models can sift through thousands of pages of regulations, contracts, and specifications in minutes, distilling key information and flagging potential compliance issues. This task, which might consume days or weeks of a human expert's time, can be significantly reduced, and allow that expert to focus on the implications and broader context that output sits within.

The second transformative aspect of AI in construction is its unparalleled ability to process and synthesize complex, multi-dimensional data. The modern construction process generates a deluge of information, from intricate digital twin models and real-time IoT sensor data to complex financial projections and multifaceted risk assessments. AI systems can ingest this vast ocean of data, discover the patterns and processes that lie dormant under the surface, and then dredge these up in ways that are actionable and intuitive to human decision-makers.

Here AI enables and empowers humans to make better informed decisions in the face of complexity, working on the value-adding refinement. Accountability and ownership are not taken away, rather the human-in-the-loop is empowered and augmented and is better equipped to make complex and challenging decisions. This capability enables a new paradigm of data-driven decision-making across the entire construction lifecycle.

In the conceptual phase, generative AI can rapidly explore thousands of design possibilities, optimising for a complex interplay of factors including energy efficiency, cost, structural integrity, and even aesthetic appeal. In project management, sophisticated predictive models can forecast potential delays or cost overruns with unprecedented accuracy, allowing for proactive mitigation strategies. And in the delivery phase, AI can intersect with the field of robotics, effectively transforming the way we affect the physical world. On-site workers can leverage robotic systems powered by AI to enhance safety, precision, materials handling, environmental impacts, logistics and program challenges. As these integrated technologies mature, on-site activities may look profoundly different in an AI and robotics-enabled future.

The catalyst

Perhaps the most exciting prospect is AI's potential to catalyse a renaissance of creativity and innovation in construction. By doing the heavy lifting of routine tasks and providing deep analytical insights that couldn’t have been uncovered otherwise, AI may free human workers to engage more fully with the aspects of construction that require uniquely human qualities, such as imagination, emotional intelligence, and complex problem-solving. AI-based tools can serve not just as labour-saving devices, but as creative partners and even colleagues, suggesting novel solutions or materials that push the boundaries of what's possible in construction.

By embracing AI thoughtfully and strategically, we can usher in a new era of construction; one where our built environment is created more productively and creatively than ever before. In doing so, we honour the essential human drive to build and evolve our environment. But the integration of AI into construction transcends our current desire for productivity; it can also empower our society to build the future we envision, to shape our world in ways that reflect our highest ideals and aspirations.