Why you should start treating AI like a new colleague

That was the consensus from a group of experts in the field of digital construction technology at a recent conference held by Nima (formerly the UK BIM Alliance).

Discussing AI for the built and managed environment, the panel agreed that despite considerations around a regulatory framework for the technology, ethics, and its impact on human employment (see more below), AI adoption is already starting to move at pace across all industries, including construction.

“The train is leaving the station. I wouldn’t worry about your job – AI isn’t going to take your job. Someone using AI is going to take your job. Get on it and start getting your hands dirty. Sit down with a few people in your company and put together an AI strategy,” advised René Morkos, founder and CEO of Alice Technologies. His company has developed a platform that allows construction professionals to upload their existing construction schedule before using AI to simulate a range of different scenarios to come up with what it claims are the most efficient construction paths.

Morkos said that technology had finally caught up to the point where it could handle the complex and often unstructured data involved in construction projects and that the technology would continue to evolve into what he called “composite AI”.

“You are going to start seeing the interplay between deterministic systems that are really good at solving math and engineering problems and neural networks to scour large information sets and summarise them for you. It will take 2-4 years to start seeing that sort of thing coming up.”

This would allow access to information about projects and decisions based on that information to speed up from a process taking weeks to just a few minutes, he said.

Murillo Piazzi, senior digital consultant at the BIM Academy, said that construction professionals would need to start thinking about AI as they would a new colleague. “We need to train the AI to increase our capacity to finish tasks with the level of quality that we want. And if we have a new colleague at work who needs to be introduced to some of our tasks then initially this may take up some of our time but in the long term it will be time well spent because it will be able to share the workload,” he said.

But Dr Noha Saleeb, associate professor creative technology/digital creativity, Middlesex University reminded construction professionals that they needed to know what they wanted out of AI before they started trying to use it.

“Before embarking on using AI blindly, perform a gap analysis that involves at least three main things: What information do you want out of AI? What resources do you currently have and what do you need to implement AI (whether that is skillsets of people, information databases, physical software and hardware)? And the third thing is whether to use bespoke or off-the-shelf solutions,” she advised.

Balance between innovation and governance

The panellists also discussed how much governance there needed to be for AI and how to strike a balance between regulation and innovation.

The European Union has already passed the Artificial Intelligence Act which aims to establish a common regulatory and legal framework for AI, while an international standard, ISO 42001 has been established to set our requirements for establishing, implementing, maintaining and improving artificial intelligence management systems.

James Chambers, director, global industry development in the Build and Construct division of software company Nemetschek, which owns brands including Bluebeam, Graphisoft and Solibri, said, “It is like any other tool we implement. There has to be a framework and guidelines and that is accelerated by the fact that AI is so powerful.”

But he added, “I think the opportunities for learning with AI are so vast given its capabilities and given its ability to mine data that our industry truly needs, that we shouldn’t stifle those. We should make sure we are playing within the guidelines and that we are protecting ourselves. The nice thing about AI is that you can put in those guidelines and get the outputs that you want from that.”

Good quality data and data sharing could enhance AI’s capabilities

Image: leonidkos via AdobeStock - stock.adobe.com

Nick Tune, CEO and co-founder of Optimise AI, which has developed a tool that uses digital twins and AI to minimise energy and carbon emissions from buildings, said that AI offered advantages when it comes to sorting through data generated during the construction and operation of built assets, which can sometimes be chaotic and poorly structured.

“Neural networks help use to be able to get some insights help us to get some insights from unstructured information in a way that we didn’t have before,” he said.

To further enhance AI’s capabilities, Tune advocated more data sharing. Although he acknowledged that private companies in the construction and real estate sectors are often reluctant to release their data into the open.

“If we could get more datasets for the common good, that would be amazing but I see many reasons why that is difficult,” he added.

Emma Hooper, vice-chair, buildingSMART UK & Ireland emphasised that having good quality data from the outset is likely to make AI more effective.

“The conversation around AI is very much around the training sets and what data is going to be used to train the algorithms. Obviously if the data is of good quality, then you are going to get more accuracy on the other end,” she said.

Concern around jobs and ethics

Panellists did express some concern that AI could impact jobs in the sector, as well as raising issues around the ethics of using AI in the design and construction process.

Dr Saleeb asked, “If I start with the ethics, who is considered liable if something happens? What if a disaster happens as a result of AI outputs? Who is considered liable for that? That is an important legal and ethical issue we need to consider.”

As far as the risk to jobs was concerned, she contended that the technology was “not necessarily that scary” because jobs would evolve organically as AI was introduced to roles little by little, at the same time that new careers linked to the developing technology emerged. “How do you co-ordinate between AI, robotics, and big data, for example? And how do you align between AI and something like the UN’s 17 sustainable development goals, which are all pertinent to the construction industry?” she asked, suggesting that this work would require human intervention.

Chambers added: “We know that 40% of this industry is going to retire out by 2030 but this month for the first time we are seeing an influx of people entering into the architecture, engineering and construction (AEC) industry, which is very positive. Maybe AI is playing a part in that because the younger generation and being incentivised and motivated by the amazing new opportunities that are being created.

“Let’s not forget that this industry that I love has got a history of doing things in a bad way: the way we contract things, the bureaucratic processes we have for site inspection and sign off. To be able to allow those really skilled people to focus on the tasks that they have so much knowledge about and not do all these bureaucratic tasks that cost a company a lot of money, to me is an amazing thing and a great opportunity. So sure, there are concerns but there are positives too.”