Unpacking the biggest questions around AI adoption

Listen to Darren Martin’s interview on revolutionizing the engineering industry with AI

Artificial intelligence stands to revolutionize the world.

The Architecture Engineering and Construction (AEC) industry has historically been one of the slowest to adopt new technology. A 2016 McKinsey report found that the industry was one of the least digitized in the world, sitting just above agriculture and hunting. Things have changed since then, and organizations are increasingly embracing the value and efficiencies technology can provide, but we still have a long way to go.

Nonetheless, AI is predicted to transform the industry. Gartner has predicted that AI will be able to complete 80% of project management work by 2030, giving those companies that can adapt and leverage it an enormous competitive edge. Global investment in infrastructure is rising, with $130 trillion predicted to flow into decarbonization and infrastructure renewal projects by 2027. And investment in AEC tech is also rising fast, with $50 billion invested between 2020 and 2022, an 85% increase on the previous three years.

So, with change on the horizon, what questions do we need to answer to ensure we can make the most of this boom and adopt AI as effectively as possible?

Let’s break it down.

Where can AI really add value?

The first challenge in applying AI, is working out which tasks AI could really add value on. This means identifying all the stages across the value chain where AI could provide a significant advantage, and assessing how practical this will be to deliver.

This also means assessing whether AI is the only or most cost-effective solution, or whether the inefficiencies you discover could also be solved by other means. For example, in the AEC market many simpler challenges can be tackled with automation tools or existing software packages (some of which may already have AI embedded in them by default, such as Microsoft CoPilot).

At the end of the day, we need to make sure we are selecting the best tool for the job, even if the best tool is sometimes not quite as flashy.

How do you pick the best supplier?

Once you’ve decided to use AI, you need to decide on a supplier. And this can lead to a lot more questions.

First up there are the classic challenges. Will they also be selling to your competitors? How much will it cost to use? Does the tool require special training? Do you use a start-up?

And if you do use a start-up, there are a variety of extra questions you will need to be sure about too. Do you understand where they are in their funding cycles and how this might impact you? Do they have an exit plan that might impact delivery? What happens if you embed the tool across your business and then they go bust?

Some organizations will choose to develop in-house, but that comes with its own serious risks. After all, just as farmers don’t tend to build their own tractors, building your own AI tools can be a very expensive and impractical distraction from the everyday business of the organization (or farm).

How can we bring people on board?

Another major challenge of AI adoption is that while many people are excited about the possibilities of AI, others may be hostile to the concept.

There is a fair amount of reluctance and pushback around AI, particularly where people feel that it is a threat to their jobs. This can create resistance against using AI resources. In industries such as AEC, there may also be legitimate fears about the fitness of computer software to carry out safety-critical tasks or make higher-impact decisions.

Some of this hesitancy will come from media coverage of AI, but it is often exacerbated by a lack of information or leadership around the topic of AI within an organizations.

How do we find people with the right skills?

AI will be a vital tool for plugging skills gaps – increasing the capacity of skilled personnel and completing low-value tasks. But in order to implement it, we need more people with expertise in data management and AI. Unfortunately, people with these skills are in short supply and high demand in today’s market.

As time goes on we expect to see an increase in people with the right qualifications, as schools and universities adapt to the changing demands of the modern world. But in the short term we also need to see a significant upskilling of the existing workforce. In fact, according to research by the World Economic Forum, executives estimate that 40% of their workforce will need to reskill in the next 3 years to ensure we can effectively utilize AI tools.

How do we make sure we are using AI ethically?

Done wrong, AI has a lot of potential for danger. And until regulation can catch up the onus will be on organizations to ensure they are using AI in an ethical manner consistent with company values.

I also think it’s worth going a step further than that and spending some serious time thinking about how to make sure your business’s use of AI is not just consistent with company values, but actively supports them. Ultimately, we need to hold our use of this software to the same ethical standards we do our people, and it’s worth giving thought to how AI can be used not only to improve client outcomes, but to support safety, collaboration, and integrity.

How can we monetize AI?

The AEC industry still broadly operates on a model where organizations charge for time and materials as opposed to outcomes. This makes adopting AI challenging, as it could easily cannibalize a business’s earnings by reducing the time they spend on a job. And on the flipside clients are unlikely to want to pay high hourly rates for tasks which are actually being accomplished in a few minutes by AI.

We need to make structural changes to the way our services are assessed and valued. This will mean moving away from time-based pricing and focusing on outcomes to ensure that everyone can capture the value that AI offers.

Darren Martin is the Chief Digital Officer at AtkinsRéalis, a world-leading design, engineering and project management organization. Darren will be discussing ‘Why a different approach is needed to monetize the value of generative AI capabilities’ on the Main Stage at the AI World Summit Americas at 9:40 on the 25 April.

Experts from AtkinsRéalis will also be running a workshop on the ethics of GenAI adoption on the 24 April at 11:25 called ‘AI – making the right decisions for your organization.’