How AI Is Changing Construction | Digital Twins, Data & the Future of Engineering

AI is already changing the way construction and engineering work gets done. The question is no longer whether the technology is coming, but how the industry uses it without losing the experience and judgement that still sit with the people doing the work.

On the UK Construction Podcast, host Jimmy Webb speaks with Nathan Marsh about where AI is making a difference across complex projects — from analysing huge volumes of project data to spotting risks, supply-chain problems and missed connections before they become expensive issues.

Nathan Marsh is Regional Executive for Europe, Middle East and Africa at Bentley Systems. With engineering and contractor shortages putting more pressure on an already stretched industry, he explains why he sees AI as an assistant rather than a replacement for people.

He also looks at digital twins, systems-based thinking and the need to treat roads, rail, water and power as connected parts of the same system.

The discussion moves into real-world examples, including the Everton Stadium and Elizabeth Line, and how technology helped teams rehearse construction, manage logistics and bring different organisations together around a shared project.

For Nathan, the real opportunity isn't simply having better technology. It's making that technology easier for people to use, giving teams better information when they need it, and making sure human judgement remains at the heart of important decisions.

The result is a fascinating look at where construction technology is heading — and what that could mean for the people delivering the industry's biggest and most complex projects.

🌐 Learn more about Bentley Systems: https://www.bentley.com/
Nathan Marsh: https://www.linkedin.com/in/nathan-marsh-6320539/

___________________________________________

🚀 Watch, Like & Subscribe to our YT channel for more episodes: https://www.youtube.com/@UK_Construction_Blog

🔔 Turn on notifications so you never miss an episode

Thank you to our sponsor for this episode, WebMotion Media:

https://web-motion.co.uk/

Special shoutout to our dynamic host Jimmy Webb of Construction Cogs: https://constructioncogs.com

____________________________________________

Follow us for other industry updates, thought leadership, house building/DIY tips and the occasional crossword puzzle!

► Website: https://ukconstructionblog.co.uk

► LinkedIn: https://www.linkedin.com/company/uk-construction-blog

► Instagram: https://www.instagram.com/uk.construction.blog

► X: https://x.com/UKCBlog

► Facebook: https://www.facebook.com/UK.Construction.Blog

► Threads: https://www.threads.net/@uk.construction.blog

► Bluesky: https://bsky.app/profile/ukconstructionblog.bsky.social

Transcript

From groundbreaking projects to game changing innovations. The UK construction podcast brings you face to face with the industry's brightest minds and boldest thinkers.

Each episode features candid conversations with construction leaders, architects, engineers, and onsite experts who share their hard won insights and behind the scenes perspectives. We cut through the noise to deliver actionable intelligence on market trends, emerging technologies, and the forces shaping British building.

JIMMY:
Nathan, thanks very much for joining us. Thanks for your time this Sunday. Hope you're well.

NATHAN:
It's a pleasure. Yeah, very well. Always great to see you and delighted to be here.

Looking forward to the conversation.

JIMMY:
So am I, so am I. So let's start then. Can you start by telling us a little bit about yourself, who you are, what you do at Bentley Systems and what the company does?

NATHAN:
So I'm Nathan Marsh. I'm the regional executive for Europe, Middle East and Africa at Bentley Systems. That's about one third of our business territorially and about one third from a kind of revenue perspective.

I've been at Bentley for two and a half years. I've been a customer of Bentley's for about eight years when I was at Costain, Atkins and other engineering consultancies. So I know the organisation well, know the sector well and yeah, delighted to be here and delighted to represent Bentley, but also give you a perspective from my career journey as well on our brilliant industry.

JIMMY:
Great stuff, great stuff. So we had a little chat before to prep for the call, to prep for the interview. You said to me that the more interesting point is not how AI code works, but how AI and better data change the way we plan and manage complex infrastructure.

We'll talk about a couple of case studies shortly, but first, what's the biggest shift you're seeing?

NATHAN:
So I think you're right. I mean, it is quite interesting how the code is developed and how it's engineered now. In some cases, it's kind of self-building and self-improving.

However, it's what we do with it that makes the difference. And we're seeing it accelerate the pace of analysis of data, the wealth of project data we might have or subsurface data, tunnelling data, bridge data, risk data. It's a lot of work to do.

And thinking back to my time working in PMOs and construction sites, that was a lot of heavy lifting. Now with an AI agent and automation, it can cut through that data, organise it for us, pick out trends, pick out anomalies. What we shouldn't do is ask it to do the job for us in terms of decision-making, but it does do a lot of the heavy lifting in a more automated, faster, and I think a more structured manner than even the best PMO leaders.

So I'm really excited that AI is here to assist how our leaders run PMOs and construction sites.

JIMMY:
Yes, yes. Sorry, I've jumped ahead a little bit there. You explained what you do at Bentley, but just summarise what Bentley does as a company.

NATHAN:
Yeah, no, that's fine. So Bentley is a software business. We focus, which I think is really helpful, just on one sector, infrastructure.

So within our sector, our software covers design, engineering, construction, and asset management of infrastructure assets across the globe. We're 41 years old. We were privately held, set up by the Bentley brothers, and still are non-exec directors.

We're now listed on the NASDAQ. A very fast-paced and exciting and successful place to be.

JIMMY:
Good, great stuff. Thanks for that. It feels like infrastructure is becoming too complex to manage it the old way, and at the same time, demand is kind of outpacing engineering capacity.

Where is that demand coming from?

NATHAN:
The demand comes from a number of areas, and in no particular order. Most, but not all, most of the world's infrastructure, particularly in the Western world, is quite old. And so ageing infrastructure is a big demand.

That means either replacement, it means upgrade and maintenance, and in some cases, we're looking at new build as well, to sit alongside ageing infrastructure. Secondly, climate change. Whether it's temperature, wind, rain, etc., the climate is going through an era of flux. So our infrastructure gets impacted by that, and our supply chain does. So we need to consider that as well. Thirdly, population change and demographic shift.

There's never been as many people in the world as there are today, and population change is moving as well in terms of demographics, age, location. And that puts demands on infrastructure, and infrastructure has got to think ahead. And then lastly, geopolitics.

We have conflicts, we have a change in demographic, changes in governments, etc., and all of those geopolitical and economic factors also influence infrastructure. So it's sitting in the middle of these headwinds and tailwinds. It's quite an exciting place to be, but it is quite demanding for infrastructure, which is why we're so busy.

JIMMY:
Yes, interesting points there. We'll touch on some of those. Yeah, the demand.

So is the problem that we haven't got enough engineers to deal with the demand, or that the engineers we do have are being slowed down by poor information and disconnected systems?

NATHAN:
I think it's a bit of both. Whilst those demands mean that, frankly, there's more work coming the infrastructure sector's way, which on the one hand is good, there is a shortage of engineers and contractors globally. That's an established reality within our sector.

So we can either pause some schemes, or more excitingly, I think, look at how software and technology can help us. And it helps us in many aspects of our lives anyway, doing things that we might have done manually a number of years ago. So I welcome the fact that software, responsibly designed and thoughtfully used, can sit alongside a workforce that isn't quite big enough yet.

That software makes up for that gap, and it makes up for it in terms of the hours needed to do the work, the computational and brainpower, but ultimately leaving the decision making and the human judgement to where it should be, which is in the hands of humans. But I think it's a welcome addition to a constrained sector at the moment.

JIMMY:
Yeah, there is a bit of a... So the industry, I mean, AI is here, isn't it? It's coming in interventions, but it is already here.

But there is a bit of a fear element to it. It's kind of a lack of education element to it in construction, but also a little bit of fear. So people are not really on board with it as much as they should be.

And you get the thing where, you know, people think, oh, this is going to take the engineers' jobs and all the rest of it. But I think it's important to know that it's not really, as you say, it is an aid to make you do those jobs easier and more efficiently.

NATHAN:
I think that as a species, anything that's new and exciting and change sometimes causes us, including myself, to raise an eyebrow. But that should be one of curiosity. And I'm really excited that when it's responsibly coded, judiciously deployed, and really responsibly used and governed, it can be a phenomenal help to our brilliant people.

And I use it most days now in my course of work, and I did when I was a contractor and an engineer. So I welcome it, so long as we go into it eyes wide open, intentionally, responsibly. And like I said, it's got good governance around it, good controls around it.

It truly is a great assistant for us all.

JIMMY:
How do we, yeah, how do we govern it then? And how do we, how do we make sure it's trustworthy and that it is coded well?

NATHAN:
Well, we can look into the code, understand the provenance of the code, who coded it initially, where, when, what was the kind of culture and mindset around? Are there any implicit biases built into the code? These are really responsible questions to ask.

We can look at how insurance markets treat it from a risk perspective. If we're going to make decisions, partly informed by what the code is supporting us to do, we need to be really clear where the liability for that decision sits. That's just good governance and being a good responsible steward.

And also understanding how investors take it as well. They're viewing AI as a really exciting addition to our sector, speeding up design, engineering, construction, optimising construction, and of course, asset management. But they do want to know that it's the right stuff and being used in the right way.

So I welcome governance. We mustn't over-govern it because we could suffocate it. So I think a responsible and flexible set of principles, standards, rules, accreditations, like most new arrivals to our sector, would be welcome to keep us all in good order.

JIMMY:
Absolutely. Yeah, yeah. I get that.

You mentioned system-based thinking in our pre-call, in our prep call for the interview. Is that basically about stopping us looking at roads, rail, water, power, and construction as separate entities and seeing them as one connected system?

NATHAN:
One of my favourite things, systems-based approaches. I think it's two things really. So in my opinion, and I saw this on the BBC a couple of weeks ago talking about this, BBC News, we saw in the UK higher than normal temperatures.

That impacted the rail network, which also then, because of some train delays, caused a pressure on the emergency services. They responded to stuck trains and passengers on them in heat. It also impacted our road network, and we saw delays, congestion, and sadly, an increase in incidents and accidents on the highway.

That also constrained our precious emergency services. The heat also, in some regions, caused a water shortage. So as I'm talking, you can see that firstly, our infrastructure assets that previously have been viewed quite discreet are an integrated and connected system.

You can also see, if you take a city like London or Manchester or Birmingham, water, energy, transport are all flowing at the same time through the same postcodes. It's a system by default and by design, and I think we're catching up as a sector to treat it as that. The second angle, I think, is on a major project, and having run and worked on major projects in the past, I've tried to bring that kind of integrated systems-based thinking to them, and I think better things happen when you do.

You've got a supply chain, you've got different contractors, you've got an ecosystem going right from the tier ones down to the trades, you've got design or even pre-design, so subsurface civils, DCO, design, engineering, construction, and then, of course, asset handling and management. Our sector, as I think we both know, in the past has been quite discreet and quite siloed and separate, and I'm part of that as well. When we stand back and look at these big projects, rather like infrastructure, they flow, the data flows, the risk flows, the opportunity flows across those different phases and stages and the points in the supply chain, so now I think the technology is there where we can truly manage all the complex components in our infrastructure assets and in a project as a system, and I'm really super excited about that really coming to its fall in 2026.

JIMMY:
Brilliant, love it. That's interesting, the BBC, talk to me about that. Can you talk about that?

NATHAN:
Yeah, of course, it was on BBC News about two weeks ago, two Fridays ago, and I had a glorious four minutes of fame. I'm sure it's a once-in-a-lifetime moment for someone like me, but I was asked to talk about infrastructure's connectivity and how, in that case, with the very hot temperatures we had a few weeks ago, how rail and roads and our emergency services and our water and power networks were all independently facing pressure, but all caused by the same thing, that spike in temperature. So I think it's responsible for us to view it as a connected system, and yes, I can't speak on behalf of the viewers, but I certainly enjoyed my time there, and I hope it went well.

JIMMY:
Yeah, I'm sure. Just quickly, I just want to interrupt the interview. I'm not sure if it's my end.

I think I'm okay, but it keeps glitching every now and again. Am I glitching at all?

NATHAN:
No, you are crystal clear that you just glitched there. Let me check my Wi-Fi. That's all looking okay.

Do you want me to go over anything again? Do you want me to repeat?

JIMMY:
No, I think that'll be fine. Let me just see. I'll keep an eye on it.

I've got you now. Yeah, so I think it's my end, actually. I'm in a venue that hasn't got very good Wi-Fi, so I see data from my laptop.

So we'll just go with it.

NATHAN:
Okay.

JIMMY:
Right. Yes.

NATHAN:
I've got two teenage daughters at home. I just told them to get off the Wi-Fi, so goodness knows what they're doing. I hope it helps.

JIMMY:
Oh, fair enough. Okay. Let's hope it does help.

Yeah. Right. Yeah.

So you spoke about the ageing system earlier. Oh, actually, quickly. Yeah.

So you talk about climate. So we kind of have a bit of a problem in this country, don't we? So that when we hit the extremes, either hot or cold, the country does kind of come to a standstill, doesn't it?

You know, the well, like you say, the well of loads of holdups, the roads, people have accidents and the roads, the potholes, everything comes from all the water. It does seem to me that we're not really prepared for these extremities.

NATHAN:
Great question, and quite literally a hot topic. Look, our infrastructure has been entirely appropriately designed for our climate. But the climate is changing, and we can learn a lot from Asia and the Middle East and Africa.

Within my regions, I visit them a lot. However, they do not build their infrastructure for our climate. So if we were to directly import rail in Qatar, or South Africa, to Europe and to the UK, there's no guarantee that approach would work.

So we've got to consider what they do for extremely hot temperatures. But also recognise that, as you said, sometimes we get below zero in the winter. So we kind of got to do that.

And it's really hard. So a lot of respect for national highways, network rail, we're trying to wrestle with these extremes. But it is something that I think gets due attention, and should get even more attention as climate change becomes more persistent, and we face more extremes and a rather arid climate impact that is coming upon us, as we saw a few weeks ago.

JIMMY:
Yeah, yeah, that's a good point. So that leads me on to this. So are we building infrastructure with enough thought about the climate it will face in 20, 30 or 50 years?

NATHAN:
I believe we are, but I believe we're starting to do it. And there are schemes that are mid-flight, that I think probably can adopt and consider that as they start to conclude the construction phase. But in a business case for a scheme that's imminent or hasn't broken ground yet, I would say it's a prerequisite.

And I would like to see either as a regulator, or an investor, or an asset owner, I'd want to understand at least that you're considering climate extremities, but also then how you're managing and that could be a change in components, a change in engineering plans, a change in design, a change in construction methods. So I think it's something that Bentley would take very seriously. So our software can support planning for the extremities, but also managing the asset typically in a more in a more moderate range.

But I think we're starting to do it more intentionally, which I think is really welcome.

JIMMY:
Yeah, yeah, I agree. I agree. Let's move on to digital twins.

I know this is another area that you're quite passionate about. Can you explain what they are?

NATHAN:
Yes. So digital twin, I think, goodness me, they've got a number of different definitions. But the one I think that's most established in our sector, it's a digital representation of an asset.

What's really exciting is that a digital twin, you start building it in the design phase. So you lay out what this asset is going to look like, and then the path to getting it. And that can be project schedules, planning perspective, risk logs, how extended the supply chain, how fragile the supply chain is, which components come from which supplier, and in which order are they assembled.

That's all really important information for maybe 5, 10, 15 years down the line when the assets handed over. And historically, in some projects, we forget to do that. Who put that bit in over there?

Who engineered that pump pass? Well, I've got a couple of PDFs held on a hard drive somewhere or a memory stick. All of that goes, a living digital twin, in our opinion, starts at the earliest stage of the project.

And it builds and it's a repository for all the planning, investment, supply chain, design, engineering, construction data, so that when you hand the asset over, you get two sets of keys. You get the keys to the asset itself, to the new owner and testing, and the ribbon is cut. And then you get the digital keys.

So tell me that technology and data story right from the beginning. And it really is a once-in-a-lifetime moment. And I'm really excited that digital twins that live the full duration of the project have such rich and well-organised data, that whoever the asset owner is when they get them, they're going to really benefit from a really solid footprint, legacy, and heritage and journey of every element of that project.

I know that if I was managing an asset, I wouldn't want to travel much further without it. So I really welcome digital twins.

JIMMY:
Yeah, it does sound great. And it wouldn't just magically create this data, that will come from areas like surveys, design models, I think you mentioned this earlier, ground data, asset records, environmental information, and construction planning, right?

NATHAN:
Absolutely. Site safety, drone surveys, which gangs, which contractors are in the fourth and fifth tier. I want to know all this stuff.

Because in 30 years' time when that asset might need an overhaul or there's an issue with the asset, I just need to know the provenance and heritage. It'll help me make the right decisions. And actually with AI within the digital twin, it can hopefully aim off and help predict what issues might happen, interrogate the data, and present me a solution or a couple of options as a human to use my judgement on.

That's very different than trawling through old PDFs and in some cases, A0 schemes and plans, which we've all done. So yeah, I think life's going to get better with them.

JIMMY:
Yes, yes, absolutely. It leads me on to a couple of the case studies. So, Bentley, you guys sent me a couple of case studies that you've worked with directly.

Everton Stadium, that was a big project. So this caught my eye because I work for Langer Rook. I actually work for Select Plant High, which is a subdivision of Langer Rook.

And I know some of the crane operators that were on that project. On Bentley's side, what was the main challenge on that job?

NATHAN:
I mean, Langer Rook are a really progressive tier one contractor, an organisation I hold in very high regard. So the Everton Stadium was about half a billion pound stadium, state of the art, it's how I opened the Hill Dickinson Stadium up at Everton. And it was really complex engineering in a very constrained environment, because around it, you had nature and wildlife, you had open waterways, and of course, you had residential and retail.

And then slap bang in the middle, you're putting a massive construction exercise. So there were 6000 components were manufactured by Langer Rook off site. So taken away, and then delivered in and delivered in at quieter times.

So that meant that the workforce that we estimated was about 3000 was then managed at 1200 at peak. So instantly, you got less than half the amount of human traffic, parking, congestion on a construction site. It was a digital build 100% digital build, which meant that that stadium had the digital twin right from the start.

In the construction phase, I know the planner quite well, the senior planner, he built that stadium dozens of times before they broke ground. And every time they tightened and optimised and found efficiencies. So really good example.

Also, the marine challenges, there were some historical docks that had to be reclaimed and technology and subsurface scanning, allow them to preserve that important heritage in the Northwest, and also build the scheme around it. So it's a really good example of technology, helping out and reducing time, reducing footfall, which makes for a safer construction site, reducing disruption, support supporting local heritage, and frankly, delivering a project in good order on time and just under budget. It's a great case study.

And now a good stadium.

JIMMY:
Yeah, yeah, yeah, yeah, yeah. On on what was potentially a logistical nightmare. Yeah, that's right.

NATHAN:
Yeah. And of course, it's very visible. Yes, the good people of Everton are watching a stadium with a close eye, the media's all around it.

So not only is that exercise under scrutiny, but you know, in some cases, they might be waiting for something to happen. But no, that ran smoothly. And it's a case study we're really proud to be part of.

JIMMY:
Brilliant. I'll bet I'll bet. The case study, it talks about the logistical constraints and a 4D digital construction approach.

I'm not going to lie, I'll tell you the truth. We're not when thinking about 4D initially, I thought imagine something like an immersive VR headset, for example, where people could walk through the project stages, personally, but that fourth dimension is more about time, isn't it?

NATHAN:
Absolutely. Yeah, it's time and schedule. And this is where I mentioned earlier, the senior planner, this is where he really saw the benefits.

So Synchro is a Bentley application. And Synchro 4D absolutely does that. It brings time schedule alongside cost into the planning.

And it brings with it automation. So it takes a static blueprint, a static plan. You mentioned earlier, we've all been pouring out the Gantt charts in the past.

And this thing was not only live, but it was able, they were able to rehearse the construction exercise a number of times, tightening it and gaining efficiencies. And I went up a few times to visit the site. You had a digital PMO that had the software and code version of what should be happening on the ground.

And you looked out of the window in the planning office and saw what was actually on the ground. So you could literally compare day by day, where you are versus where you should be. And if you're not quite there or ahead of plan, you can make some adjustments as well.

It's a great case study of a living plan.

JIMMY:
Yeah, brilliant, brilliant. And this, I mean, additionally, this system, I imagine it's great for winning work, isn't it? I mean, you mentioned Gantt charts.

So it's better to turn up with this system than with drawings and Gantt charts. And you can actually show how the project will be built visually.

NATHAN:
Yeah, I think so. I mean, instead of a complex, disconnected text-based spreadsheet or Gantt chart, and we've all been there, by the way, site managers will now click through a chronological timeline. So the digital model physically pieces itself together on screen, step by step.

Years ago, it was post-it notes on the wall and moving things around a flowchart and a Gantt chart. So I think it's really welcome.

JIMMY:
Yeah, I think this is useful on all scales, really. I mean, that's quite a large-scale project, a very large-scale project. But even like your one-man bands, I think this is useful for, isn't it?

You know, someone's going to a house or whatever and you're pricing work, the owner, they want to know about amendments, they want to know how this is changing, how is this going to judge that price, the timescale on that. So you can just look at the system rather than just WhatsAppping and emailing and texting and losing track. Just have a system in place where you can just track everything.

NATHAN:
Well, it's certainly one source of the truth. And according to your access privileges, it's access all around for different, you know, the lead contractor will get a particular view. The project planner will get a view.

The project director gets another view. Stakeholders and investors might get a view. Environment agency colleagues will get a perspective.

But all from a single source of the truth. I mean, designers, engineers, contractors, the full supply chain. And as you said, right down to the trades to make sure that my time isn't wasted when I turn up with my gang on site.

You know, my margins might be quite tight, but I want to benefit from the knowledge that is held in this thing. So it's a, it's one portion of the truth, access all round, subject to your privileges, and it's constantly live and updated. God, sounds, it's going to make a hard job a bit easier, isn't it?

JIMMY:
Absolutely. Absolutely. Let's talk about the Elizabeth Line case study now.

So this is kind of related to what you were speaking about earlier with the race. It's a, it's very good systems example. It's not just what's built, it's how people use it once it's open, right?

NATHAN:
That's right. I mean, what a lesson, something for London to be proud of. I use it quite frequently.

JIMMY:
Yeah.

NATHAN:
It's a, it's a beautiful network. The engineering was highly complex. And I remember when I was a CDO at Costain, the engineering complexity and those big boxes that you put in the, the new stations in was profoundly complex.

But again, a systems based approach was taken during construction and design and engineering. And of course, getting the, getting the asset up and running, getting it handed over and now working beautifully. It's a great example of really well synchronised schedules, plans, different trades, and quite frankly, rather like another major projects in the UK, competing organisations coming together and having a, having a collaboration agreement and taking a systems based approach on a project really works.

Just look what's happened. That thing runs seamlessly. It's something to be proud of.

And we need to look at that as a great case study.

JIMMY:
Yeah. Can you, can you just touch, expand on that a little bit, the competing organisations? Because, because that hits home because obviously I've worked on quite big projects, but I tend to find on sometimes there's a little bit of political element to it.

So you've, you've got the contractor, you've got the subcontractors, there's a little bit of tit for tat. And I think if you can actually, you know, get over your differences, you know, just actually work together on things properly, it'll be a lot easier.

NATHAN:
I think it's about, it's a difference between, or the blend of IQ and EQ. The IQ in it, we all know the organisations of the big engineering firms, medium and small engineering firms are all quite rightly independent, competitive and successful. Same for contractors.

However, on projects like Elizabeth Line, High Speed 2, Thames Tideway, they've got to work together. No one organisation wants to take on the risk. So we share it and we share it between phases, regions, tunnelling, bridges, etc.

And so these collaboration agreements, I think are super, super important. And they stop organisations. And I've been in, in organisations that have competed.

I mean, you get onto the project and you work together, you have a project charter or a collaboration charter. You sometimes see even small things like not using your corporate logos. And you have a project logo and a project brand, and you're part of that project rather than working for a particular contractor.

Of course, you'll always be working for the contractor from a pay and rations perspective. But having these EQ and IQ sort of elements covered is really important. But it's the EQ, I think that willingness to collaborate, that really defines a partnership model.

And that also then brings on our sector together like a system. So if we act like a system of integrated people, Working with a common goal to deliver that project. Then we used system-based approaches in the delivery of the project. Good outcomes are going to happen and that asset will be delivered as a well-integrated system by a consortia that can be proud of it.

JIMMY:
Yeah, absolutely, absolutely. So you've got kind of stigmas about the private sector and the public sector. I think more now they are coming together to deliver these big projects and they have to at the end of the day, but you have no choice.

NATHAN:
Yeah, I quite agree. I welcome the collaboration agreements and you see organisations certainly put down their competitive instincts for a moment and get behind the project win. I think that's good for our sector and good for our country.

JIMMY:
Yeah, definitely, definitely. So yeah, so the Elizabeth line, so that is now done using this systems-based thinking. So how much of a challenge is it now to incorporate this system into retrofitting, like you say, these aged systems that we have in place already?

That must be really challenging.

NATHAN:
Yeah, I mean, I think on the Elizabeth line, we can see it quite nicely integrated into the underground network and the overland network in London as well. Because I think taking a systems-based approach is relatively new to our sector. If we remember that rail, highways and water assets, particularly in the UK, one, two hundred years old, we've been doing this for a small number of years.

We're getting more and more proficient at it. I think we've got a little more time to tell, a little more time to run before we can say with confidence that systems-based approaches are delivering true assets that operate as a system. We're making them operate as a system because we've inherited them, and that's OK, managing the interfaces and the integrations, but designing them as a system and then launching them as a system, I think, is the next phase.

And that's super exciting. So it's a great time to be in construction with that at the fore.

JIMMY:
Yes, yes, absolutely. We touched on AI earlier and real world data. I'd like to expand on that a little bit.

I mentioned this point earlier as well. When people talk about AI in construction, a lot of people jump straight to fear that it's going to replace engineers, but it's not. When you've got digital twins, live data, complex infrastructure all coming together, what does AI actually help with?

NATHAN:
So I think AI can help with a number of things. It can certainly help with spotting those interfaces and integrations I mentioned earlier. How immediately apparent to a human eye and a human mind is it that right down the supply chain or right into a seventh or eighth phase of a construction project, there might be a number of items, and they could be commodities, components, suppliers that need to come together.

We have to manually interrogate that. And I know that because I've tried it myself. Or you let well-trained AI, let it loose on well-structured data and say, find me the interfaces and integrations.

And it will scan project data, scan a schedule and say, hold on, these two things are not connected and they should be. That is so helpful. So we don't have the oh no moment when we realise that a load of heavy freight is turning up on a Monday and we're expecting it on a Sunday.

And that does happen. Have we got the right tunnel boring machines? Those are expensive assets.

Are they turning up at the right postcode or the right grid on the right day at the right hour? They need to, or AI can help us with this. That's the first thing it can do.

It brings precision to our planning and I welcome that. Secondly, it can spot trends. It can spot trends in risks.

So are we overloading the supply chain? Given the geopolitical issues that face our world, sadly, at the moment, are we happy, for example, with commodities, essential commodities like steel coming through a very constrained and potentially troubled supply route? Well, AI can spot that and say, did you know that that contractor uses that supply route for that commodity?

It might be okay. But look at the geopolitical risk over there. I'd rather know.

I'd rather know. So it's got that scanning capability for me. It's also good at calculating, calculating risk profile, calculating budget overspend, and hour by hour recalculating where we can redeploy project capital back into the project bank account.

If we're reducing the risk on a phase of the project, because AI brings us more knowledge and we can aim off and prevent a risk happening, the budget associated with managing that risk, in theory, can get returned to the project bank account. We could use that elsewhere. So we're enriching the project.

Just those three examples, I think, make life a lot more enjoyable on a project.

JIMMY:
Yes, enjoyable. Agreed. So I can give an example related to one of those examples you spoke about, like steel coming in way ahead of time.

So I help build high-rise buildings. Some of them are on really small footprints. So you have to manage that small footprint.

So sometimes we've had like, say we're building slabs with rebar. Sometimes we've had like lorry loads of rebar come in that are for seven stories above, seven floors above what you're working on. So then that's got to be stored somewhere.

You're putting stuff on top of it, and then you've got to take stuff, that stuff off of it to get to that. And it's just a nightmare. And it does my head in for a start, being a crane operator, giving me extra work.

It's more risk because you're doing more lifts. It's just creating more mess, slowing the project down. So yeah, this is where AI helps, definitely.

So that is a perfect example right there.

NATHAN:
Yeah, I think it really does make life better, safer, faster, greener, cheaper. But just to underscore a point I made earlier, I think absolutely judgement and decision making resides with us, the humans. So you're basically saying to your AI, thank you for getting me this far.

Probably couldn't have got this far without you. I'll take it from here. And then we apply that judgement and that kind of safety critical and risk critical judgement.

I think that's really important to reinforce.

JIMMY:
Absolutely, definitely. That was actually my next question, actually, what should AI never be allowed to do without human judgement? Yeah.

Yeah, well, exactly that, human judgement. You shouldn't be allowed to do that. Bringing it back to the wider industry.

So is there a risk that large clients and major contractors move ahead with this technology while smaller firms get left behind?

NATHAN:
Well, that's a good question. I think there is a risk of that. There is a risk of that.

That said, the technology that we all use, or largely use in our daily lives is profoundly different to the technology we used, didn't even have 15 years ago. So the first point I want to make is that the technology is pretty accessible. Now, some specialist technologies are necessarily quite expensive.

Some of the planning tools, construction tools, and you have to have a certain skill to use them. And that's okay. But we need to make sure that it's accessible economically and technically.

So using natural language to go from a user, maybe a user with not much experience or training, straight into a complex engineering tool, historically, has been quite hard. Natural language makes that a lot easier. You're speaking to the application in plain English and what you want it to do.

The code and APIs then interrogate through agents. The best way to answer that question, giving access to the best applications. Some will come at a cost and some might be open source.

But I think that we all, right down to the trades and the individual trades as well, we all have a duty to embrace technology and think about, I don't necessarily need to be a super user on, say, project-wise one of our applications, but I need to know it's being used. And then my role in that might be that I might get a report or an automated update, which has made my life easier. So as you said earlier, we're not going to get the small business or the small trade turning up with the wrong equipment or on the wrong day, because that can really hurt them.

So I think it can help us all if it's used in that kind of collaborative and team spirit.

JIMMY:
Yes. Great stuff. Great stuff.

If we focus on the technology itself for a moment, what still needs to get better? Is it the AI, the digital twins, or the way all these systems talk to each other?

NATHAN:
That's a great question. I think the human interface, the human machine interface, which is a very techie way of saying how easy it is to use, that can always get better. We work with an organisation called Unreal Engine, who are a spin-out from a Hollywood studio, and Unreal makes our engineering applications, they put a front end on it, and it makes them incredibly easy to use and very visual.

So almost like watching a Hollywood movie or using gamification technologies, where in years gone by, you had to be technically highly proficient with lots of keystrokes to use some applications. They're now more intuitive, they're now more realistic, they're using a far more human and natural interface. I've talked about the voice guide, natural language, so you're telling an engineering application, move the car park 30 feet to the left, that in a planning tool.

It might not be 30 feet, it might be 27 feet, but the software knows that, so you don't have to. So it's just, I think, the human interface is really important. I think AI will be looking after itself, with that good governance around it, so I don't worry too much about the pace of AI development.

In fact, I think that's happening at a really exciting pace. How it's used is the application of software that I think our industry needs to get better at. Not using software for software's sake, not using it for everything all of the time, but working out on a major project before you get going.

What does our technology strategy look like for the project? What applications, what APIs, what systems, what data flows do we need? What don't we need?

Where do we not want to use it? I think they're really important, they are really important considerations. So that's less about the tech and that's more about how we use it and when we use it.

Let's also remember that we want that technology to leave a bit of a legacy as well. So we'll be learning with AI alongside us, how can we then take that to the next project to scale those efficiencies and frankly never make the same mistake twice.

JIMMY:
Yes, I think the human interface is a really important one actually because you don't have to do a degree to be able to learn how to actually use the AI, that's kind of counterproductive, counterintuitive in itself, isn't it?

NATHAN:
Yeah, that's right. I often say if I can get my mum to use it, we've probably got it about right, which makes it very accessible, which is really exciting.

JIMMY:
Yes, brilliant. We are done with you. You made some really, really great points there and what I normally do now, I tend to invite the guest to ask me a question.

So would you like to ask me a question? If so, do you have one at hand?

NATHAN:
I would absolutely love to. What do you think is the most exciting future application of AI? Where do you think it can help us in the near future, from your experience operating a heavy plant?

JIMMY:
Operating a heavy plant. So, yeah, I'll actually talk about something Select are doing actually. So for a plant hire company, you've obviously got lots of plants, so they're producing a telekinetic system where you can monitor the machine's usage, the power output, if it's supplied by the main power source, power output, which is great for the carbon footprint, maintenance, shutdown time, breakdown time, fuel efficiency.

So all of these things are talking to each other, so that can help them to produce better packages for their clients in the future. That's one aspect for me, particularly on tower cranes. So for me, I use, to lift loads, it uses a certain amount of oomph, a certain amount of power.

So companies, they want to know how much power we're using on this project. So they can look at that and go, okay, for this lift, massive 20-tonne lift, this is using this certain amount of power, so they can monitor that. Wind, downtime, all that sort of stuff.

Yeah, it's exciting.

NATHAN:
And that's really interesting, isn't it? And that literally comes from your seat, the kind of view from your seat, that AI could help your organisation to help Select manage its power output, which is great for the environment, great for the bank balance, can also support site safety as well, and operative safety as well. So what a great practical example.

I think it really can help us all, like I said, if it's not right. Yes, absolutely. Absolutely.

JIMMY:
And I'm saying that from a year or two ago, as a real naysayer, oh no, AI, no, can't use that, don't use that, no, and I'm converted. Nathan, it's been brilliant. Thanks so much for your time.

It's a pleasure.

NATHAN:
I really enjoyed it. I really enjoyed it. Thank you for making it very enjoyable and interesting.

JIMMY:
I'm pleased and more than happy to do it again in the future, on another topic or another topic again. Yeah, great stuff. Enjoy the rest of your Sunday.

NATHAN:
Yeah, take care. And you, all the best. See you soon.