Richard Ceeney and Kyle Sethi of Reed Smith explore the benefits of AI systems for construction
AI is fast having a revolutionary impact on a whole range of industries. Indeed, some estimates anticipate the global economy growing by as much as $15.7tn by the end of this decade due to AI, with over $6.5tn of that growth expected to come from productivity gains.
But one sector that may not, at first glance, seem like an obvious candidate for deploying AI is construction, given the industry’s historic focus on the physical aspect.
However, construction is a multifaceted process that involves far more than just the physical act of building, whether in the form of houses, bridges, data centres or skyscrapers.
AI can, and increasingly is, making contributions across the entire lifecycle of projects, offering increased efficiency and sustainability while enabling innovation.
That is particularly valuable for the construction industry, with productivity levels having remained consistently lower than the UK average while, on average, 98 out of 100 megaprojects (projects valued at over $35mn) run over budget by 30% and 77% overrun their construction programme.
That points to the extent of the challenges facing the construction industry. So how can AI help?
Applications of AI
One of the more obvious applications of AI to the construction industry is in the design and planning stage.
By way of example, rather than relying on manual processes to design a complex building model or models, construction professionals can use AI-driven software to complete the same task in a fraction of the time.
By the same token, it is possible to simulate potential issues with the design before construction begins in a level of detail that was not previously possible. The result is quicker and more informed decision-making at a lower cost.
AI can continue to offer efficiency savings into the construction phase of a project, for instance through project management software.
Returning to the physical aspect, AI can contribute to the physical act of building by powering robots that take on bricklaying or concrete pouring, or to fly the drones that conduct site inspections. Not only does this increase efficiency but it also improves safety.
Sustainable benefits
Sustainability is a key concern for the construction industry. Construction has a significant carbon footprint, which means it will also have to play a role in achieving the UK’s legally binding targets to reduce emissions by at least 100% of 1990 levels by 2050. This is putting pressure on the industry to operate more sustainably.
Fortunately, AI is well placed to support these improvements in sustainability.
For example, AI can be used to optimise material use and reduce waste, which will be crucial in reducing the industry’s carbon footprint.
AI systems can also be used to determine which building materials will be most environmentally friendly and to monitor energy consumption during and after the construction process to ensure maximum efficiency.
The growing role of 3D printing
By deploying AI and 3D printing in concert, construction professionals can use rapid prototyping and produce complex structures, while minimising material waste without compromising on the ability to create custom designs.
Another advantage offered by 3D printing that can be maximised by AI is on-site fabrication, which offers the twin benefits of reducing transportation costs and increasing efficiency. Reducing a reliance on transporting materials to the site also reduces emissions.
Where next for AI systems in construction?
Although AI is playing an increasingly prominent role across a range of different construction processes on projects all around the world, the industry is still in the early stages of integrating AI.
As a result, the full potential of AI for construction has not yet been determined and there are still a host of potential applications of AI that have not yet been deployed, with a large number of proptech companies, for instance, attempting to unlock these new uses and the accompanying benefits.
This also means that there are legal considerations for using AI that have not yet been definitively settled by either the courts or governments.
Consequently, for all of the important benefits that AI can offer to the construction industry, its use should also be approached with caution. For instance, the huge volume of data processed by AI systems could pose data security challenges.
And what if something goes wrong?
The question of who is accountable if, for instance, an AI system produces a non-functioning design or a bricklaying robot causes structural damage also needs to be considered.
There are likely to be several parties involved in AI systems (including data companies, designers, software manufacturers and end users) and attributing liability therefore for any issues arising could be very difficult in practice.
Any number of parties could be responsible when something goes wrong in a construction project.
Proving responsibility is another matter. It may not always be obvious what the root cause of any issue is: whether it is caused by AI software, manufacturing issues or human error, for example.
And, based upon experience to date, it is highly unlikely that the producers of the AI system would accept liability for the consequences of errors.
A key question, then, is who is employing the AI system. If it is the developer then it may be that nobody is liable to the developer if the AI-generated product is defective.
Construction professionals using AI systems may find themselves more exposed in projects; it is likely that their professional indemnity insurances will not cover claims arising from designs generated by an AI system, nor that latent defect insurances will cover claims based upon AI-produced design, workmanship or materials.
Where use of an AI system is mandated, consultants and contractors may seek to pass the buck to developers to account for the risk of AI errors, transferring this risk on to the client in their terms of engagement.
This, in turn, could pose a major obstacle to adopting such systems on projects as developers may be unwilling to expose themselves to these large – and potentially uninsured – risks.
Indeed, the novelty of AI technology means that the introduction of any system comes with the risk of it containing errors because it has not been properly trained, tested or it is simply new to the market.
AI systems are brilliant for identifying human errors and assist in designing and constructing safe and technically compliant buildings and structures but the very systems themselves may also fall foul of human (or system) error.
All parties involved in an AI construction project will need to consider processes to ensure that they can identify when an error has been made and a way of identifying or tracing what error has actually occurred so that it can be rectified (if, in fact, it is capable of rectification). The difficulty is that it may not be simple to troubleshoot or verify AI-generated design, or to provide effective QA/QC for AI-produced materials.
The potential damage and/or losses that may be caused by such errors could be significant and unknown.
Unless and until AI systems are proven to be reliable and defect-free, AI is likely to be more useful (and safer) in checking designs that have been generated traditionally than in creating its own designs, even though this feels like a tethering of AI’s capability akin to the requirement that the first cars be preceded by a pedestrian waving a red flag.
Still, that should not distract from the advantages that AI will be able to offer to the construction industry.
Given the numerous challenges that the construction industry faces, including in delivering projects efficiently at scale, on budget and sustainably, it is arguably particularly well-placed to harness AI to begin to tackle these challenges.
AI may not be a panacea nor will integrating AI into the industry be without challenges. But those challenges pale in comparison with the significant benefits that AI systems can offer to construction.
