EP 132

86% of workers believe they need to up-skill to cope with the changes that AI will bring. How does the RICS recommend Quantity Surveyors do this? (EP 132)



This week, Paul is joined by Andrew Knight, Global Data & Tech Lead at the Royal Institute of Chartered Surveyors (RICS), who I think everyone will have heard of. Andrew is a highly experienced technology specialist and is currently leading the program to develop, enhance and gain adoption of RICS’ Data Standards.

In today’s show, Paul and Andrew discuss how a Quantity Surveyor’s skillset is already changing regarding cost estimation, take-offs and more while focusing on Chat GPT and a recent report on AI by Boston Consulting Group.

86% of workers need to be up-skilled to cope with the changes that AI will bring; this episode is your first step.

Your free OTB download

Have you ever wished you could summarise a contract in a couple of pages for yourself or your project team?

Last week, Paul mentioned the Contract Audit document that perfectly summarises a construction contract in just 3 pages. If you’d like access to that, drop Paul a line below, and he will share it.

Want to connect with Paul?

Paul is on Linkedin here and would love to talk. You can also connect with Paul at paul@c-link.com or through, C-Link. 


C-Link is software built by Quantity Surveyors for Main Contractors. We save 600 hours of Quantity Surveying time per project in automation. We can make your QS’ so much more efficient.

Book a demo to learn more by clicking here.


Paul Heming: Hello and welcome to episode 132 of the Own the Build Podcast. With me, Paul Heming. I am going to re-share what we did last week. Have you ever wished you could summarize a contract document in as little as three or four pages for you or for your project team if you have wanted to do that, and it particularly aligns with what we discussed last week about understanding the contract? Before you start, just give me a shout. I’ve got a contract audit document, which I can share with you. You can either connect with me on LinkedIn and send me a message or email me at paul@c-link.com. Just give me a shout. I’m more than happy to share that with you. In the studio today we are rejoined by Andrew Knight, Global Data and Tech Lead at the Royal Institute of Chartered Surveyors, the RICS, who everybody knows. Andrew is a highly experienced technology specialist and is currently leading the program to develop, enhance, and gain adoption of RICSs data standards. Not only that, but he stands right on the precipice of taking the gong of the number one own the Build Podcast. He’s almost there, he is at number two at the moment. He’s just about to go to number one. Hence wise, back on the show. Andrew, I’m blabbering again. Great to job you back. How’s it going? 

Andrew Knight: Great to be here, although currently it’s a sort of record time. We’re in a fairly hot UK, aren’t we? So as usual, you need to get used to this, but no, very glad to be back and waiting for my gold disc.

Paul Heming: It’s in the post, it’s on route. Well, I’m going to send you a silver one for the moment. Let’s see. You’d never know the wind might change. So you say you’re in the UK now, you’ve just come back from Toronto, haven’t you? Tell me what were you doing there and how did the trip go?

Andrew Knight: Yeah, it was a useful trip apart from the air quality as we all know, there’s been these terrible forest fire. So actually it wasn’t the sunniest part of the world and the air quality was very kind of substandard. It wasn’t quite the kind of orange yellow pictures we’ve seen, that past had moved. That bit had moved by. But it was still not great in terms of air quality. But on a serious note, I was there speaking actually at a business valuers conference because the RICS membership includes a really broad canvas of professionals, not only in the construction and project management sector, but also in that valuation sector and not just real estate. You know, we have members in that business valuation side, and actually, I was talking about a paper, I wrote last year around automated valuation, which is an increasingly important element of the mix of people using AI and statistical analysis to look at the value of property. So a really interesting topic and one that is also being used in the business valuation space as well. So that’s why I was there, and that’s what I was speaking about.

Paul Heming: Excellent. And on environmental circumstance, obviously you’re absolutely right. We did see those almost apocalyptic video shots, photos where it was such a peculiar color, the sky, and it was just seems so hazy. Like you said it wasn’t that bad or was that slightly different area? What was it like to be in a space which was being affected so much by climate and what’s going on? I asked this because my cousin lives in Sydney, and I remember him describing to me the forest fires that obviously went on for so long. The bush fires went on for so long in Australia, and he was living in Sydney and Sydney was engulfed in a cloud. So what did it feel like?

Andrew Knight: Well, I guess for me that the impact was perhaps less strong because of that lack of color because that particular piece had moved down the east coast towards New York and DC area. And so it was a more subtle impact in a sense. There was just this lack of visibility and really the fact that you’d bump into people in the hotel and say, we’re just going to go out and check the air quality and recommendations. Not really to go outside and exercise. And clearly visibility was poor, but it didn’t have that huge impact perhaps as had been seen a few days earlier in Toronto. But it is just another reminder that for better or for worse or for worse, the effects of climate change are already here. That they have tangible effects and that air quality is, as we know, has such an effect on health depending on people who’ve got respiratory issues, it’s not just intriguing and interesting photographs. These have real world impacts on people who are vulnerable to these kind of things. So it was yet another sobering indication if we needed anymore that the effects are here and they can be very visible and newsworthy, but the more pernicious effects are there for real people who can’t go out. And so I suppose there was a certain irony in seeing people wearing masks again.

Paul Heming: It’s peculiar, isn’t it? And this might sound like an unusual thing to say, and this is obviously incorrect in many ways, but in the UK at least currently, you don’t necessarily feel the dramatic shift in climate events, if that’s not the correct way to put it. I’m talking about my cousin in Sydney, or this event happening in Canada. It is always been a little bit more mild, hasn’t it? And certainly currently it feels a bit more like that, but did it feel like a bit of a glimpse into a quite scary future?

Andrew Knight: Well, I don’t know, perhaps I’m being slightly more, not so much dramatic, but I mean, if I reflect back on last year in the southeast of England, I could see some fires that have broken out on the Surrey Hills from the top of my house. So whilst they weren’t in my back garden, I could see the fact that were fires. And in fact, there was one day I remember distinctly from an air quality perspective where, and it sounds very much like a first world problem here, but my wife said, somebody’s running a barbecue. I thought, I’m going to bring the washing in because we walked outside and it stank of burning wood, and it wasn’t a barbecue, it was the wind had blown the smoke from these fires in the Surrey Hills, and they went as far as Heathrow, and I think they caused some disruption at Heathrow. And I’m sure over the last few years, we can all remember some of the terrible flooding that’s taken place, particularly those kind of really intense storms in the winter that have just simply overwhelmed certain places in Cambria and other parts of perhaps the northwest particularly. So I would argue that actually if we reflect then even the UK is beginning to see, and there’s already seeing some of these really significant effects.

Paul Heming: Yeah, no, I think you’re absolutely right. I guess it’s the dramatic nature. I guess maybe it’s just the thing I’m associating with such large fires and uncontrollable fires, they go on for days and days, but I completely agree with you. We have gone way off tangent, just simply question… 

Andrew Knight: We have a bit, haven’t we? 

Paul Heming: But that’s really interesting. I know it gets your mind going as to where… well, as to many things really. But so grounding the conversation back in construction and we’re actually going to link back to some of the things that you just said. So talk to us briefly about your journey into the RICS and what you are now doing in the RICS.

Andrew Knight: Certainly, I will do. Yeah. I mean, I’ve been with the RICS for 12 years now, and I guess one of the beauties of working for such a broad profession, both in terms of the practitioners we have in the UK but also globally, is the experience it teaches you to really understand and get under the skin of such a wide ranging profession. So in those 12 years, I’ve worked on running teams that were helping people get into membership to access our training products. I’ve worked with our public affairs, team working particularly with the capital markets to understand the interface between lenders and funds with the profession, whether that’s through valuation or other kind of services. I’ve worked for a number of years, as you say, leading the work on looking at data standardization. And now my role is gratifyingly sort of wide ranging in the sense that it’s working within our knowledge and practice directorate, looking at thought leadership and analytics on this whole impact of data and technology, not just on the construction sector, but this broad profession that I represent. And certainly it’s been a fascinating 12 years of adding all that built environment knowledge, working with construction firms as our member firms, working with individuals in that sector across, as I say, construction and the other practice areas. And then applying my historical data and technology knowledge. It was funny, I almost do it as a party trick when I talk at conferences is to get a show of hands of people who remember what a punch card is. I asked the audience in Canada, and actually they must have been gratifyingly old. A punch card. Exactly. Yeah. Going back to the days when I learned how to code in Fortran on a mainframe and didn’t even have a green screen. We had to sort of type out our code in Fortran on punch cards.

Paul Heming: There’s a lot of confused listeners, right now. I was going straight over their heads, Andrew.

Andrew Knight: Exactly. Yeah. Well, they’ve got Google that they can look up and get an image of a punch card on Google, and they still won’t be much the wiser. But to suffice to say that I’ve always been involved in technology in some of the firms I ran in previous lives. I literally computerized from a totally manual system back in the noughties. I worked for a firm that was doing a lot of quite deep statistical work on consumer behavior and using that to what would now probably be called AI in the terms of using regression analysis to look at predictions and modeling of customers using a whole room full of Linux machines. So a long pedigree of working in data and technology and then a really good solid 12 years here at RICS, really getting under the skin of I say construction and the rest of that property lifecycle because they say our members really see property all the way from birth to death, from land, construction, brokerage, valuation financing, asset management, building operation, building surveying, end of life. And I think perhaps without prejudging the rest of our conversation, it’s really important that we have a profession that looks at assets in that whole life piece now, from where do I build, how do I minimize and maximize the positive impacts of development? How do I then construct that? Look at the issues of carbon and cost, but how do I think about the whole life of that asset and how sustainable can we have those assets? We talked about climate change and air quality and for federal for worse, as you know, that the built environment has had a huge historical impact on carbon emissions and we need to really confront that in a positive sense and do a much, much better job about managing the impact of real estate.

Paul Heming: No, I completely agree and I mean, you taught, I guess, the event that you went to in Toronto where you were talking about valuation that was not valuation in the quantity surveying sense, as I understand that, that’s valuation in the real estate sense. So I know that you deal with a broad spectrum, but only very selfish here. For our listeners, we love talking about construction, right? So a lot of quantity surveyors is, I’ve got to be selfish and champion us. We talked, if you’re listening to this now and you haven’t listened to the first episode that we did with Andrew, like I say it is one of our most popular yet really interesting. We talked about the future of the quantities surveyor, I think its number 114. So go and check that out. In that we talked about not the extinction event of AI and data and technology wiping out quantity surveyor. We talked about the importance of professional skepticism as you eloquently put it. And by the way, I’ve run away with that phrase and saying, and I say it to everyone now, I sound I’m very, very clever. So thank you for that. You talked about how quantity surveying, surveying generally, talking broadly with the RICS, isn’t going to be wiped out. In fact, data technology can actually improve a lot of the things that we’re doing. And I wanted to just drill down further into that today. Because you know, the episodes kind of ran away with us a little bit when we chatted before. So what I wanted to talk to you about today and paint the pitch, because I’ve had a lot of conversations with surveyors who think, yeah, AI data, whatever, yeah, technologies, there’s no way it wipes out QS. You need professional skepticism, I agree, but professional skepticism based on a much better platform of data tech, et cetera, to make you much more automated is what we discussed. What I want to discuss with you today, I will get there eventually, is what has already changed in the future of surveying. You were talking about valuation and AI for valuation in the context of real estate. What has already changed for quantities surveyors; benchmarking, cost estimation, BIM models, talk to us about that.

Andrew Knight: Well, I think a lot has changed and a lot hasn’t, and I’ll explain what I mean by that. There are a lot of tools out there, there are a lot of pieces of automation that look at, as you say, cost benchmarking. There are lots of extremely powerful BIM models, digital twins. I think what hasn’t changed is, is what I would use almost to phrase the kind of democratization of that, is the availability of that technology for firms of every size. And so I think in the same way that people would probably agree that the arrival of Excel has not killed off the quantity surveyor. I think we should say that the increased adoption of these tools that are really there not to replace, as you said, the professional skepticism, the judgment, the experience of quantity surveyors. But what these tools can do is stop the re-keying of information, the time spent aggregating information from different sources, the time spent trying to normalize data that you have coming up the chain from various contractors. And so it’s about saving time in arguably the least attractive parts of these jobs. Unless you particularly like just spending your time sort of massaging data and trying to get data from different sources and trying to turn a PDF into an Excel spreadsheet and try and extract information or indeed just try to record information during the life of a project. So I think it’s around looking at the right end of the telescope here and saying, well, actually, what can these tools do to remove these mundane tasks that actually suck up a lot of project time and have knock-on effects? If we haven’t got the right data at the early stage, when we get to the claim stage, we’re going to have issues because we haven’t got the right trail of information and data to understand how to resolve those. So I think it’s about treating automation and these tools as tools, just like a disto for measurement or excel for putting together some kind of spreadsheet. So I think we need to think of them as these basic tools for handling, analyzing data that saves us the hassle of having to build the model ourselves in Excel or transpose information from one document format to another.

Paul Heming: I completely agree with you and I think you, if you excel is a really tangible, easy to picture example, right? In 1985 or whenever it was that Excel came out, and obviously it wasn’t adopted instantly by the construction industry, but all those years that passed, if you think about what Excel has done versus manual plus calculator for a QS, I think how much time it has saved for you. It’s exactly the same functionality that’s going to go forward and it’s going to enrich and improve the dataset. Because Excel is far better than paper for data, but is it the optimum for us as a sector? Absolutely not, right? We’ve talked about that before, so you can really picture the change there. I guess the question is, your advice to quantity surveyors about upskilling, like how do they upskill, where do they upskill? Do they just need, you talked about the democratization of software. I worked for a smaller company, my company wouldn’t have been anywhere near this software for a long time, whereas imagine if you’re at the tier one, you’re much more likely to see it, to feel it or whatever, to feel the impact of it. How can you, practically, how can the RICS help or what can be done to upskill to take that step?

Andrew Knight: I mean, I think a part of it is awareness is education. We can do what we do with our tech partner program just to highlight the kind of tools that are out there because there are huge range of people developing and also existing tools that have actually been around quite a long time. And from that point of view, given the nature of the sector and yes, it is fragmented and many of these, the firms as you know are relatively small. It’s about also perhaps being cheeky and saying to the tier ones, well, can we see what you’re using? Can you help us educate? Can that knowledge and awareness trickle down the supply chain and help people understand what’s available out there? 

Paul Heming: But is that something the RICS champion though? Sorry to interrupt you. Because you know…

Andrew Knight: I think we can certainly talk about it and champion it and I think we have to, because I think if you look at the kind of training that’s out there, I think there is this big gap between sort of vendor specific training, if I call it that. Well, people will teach you how to use product X, Y, Z, sort of BIM modeling product X, Y, Z, and then obviously the very important fundamental training of the QS skills that will never go away in terms of the fundamentals of what a QS should know how to do. And it’s not that in some respects QS should not be taught or forget how to do the things fundamentally almost from a blank sheet of paper in the traditional way. But let’s look at automating those things. But I think it is a case of actually being curious if you’re in a small company working with your boss and saying, look, can we look at what’s out there? I’m not saying we need to buy it tomorrow, but can we look at what’s out there? Can we look at what our competitors are using? Can we look at what the tier ones are using? Is there some way we can gain knowledge and expertise from them in terms of some of these very accessible products that will, as I say, replace these mundane tasks and free up time for that quality type thing?

Paul Heming: Should QSs do you think still study and learn how to do a takeoff with a ruler or not with a ruler, with whatever should, in the future I imagine, in the quite near future, I imagine, quite a bit of that is going to be completely unnecessary. But if you go scale it right back to…

Andrew Knight: I suppose I would argue it should be unnecessary on a day-to-day basis but I think there’s an argument in every profession that we should understand how the fundamental process works to retain that wonderful phrase, professional skepticism. If you’re looking at some wonderful tool that either looks at a 2D or a 3D BIM model and does the QTO for you to be skeptical, you need to know, well, how would I have done that? Here’s this drawing. It’s telling me, I need quantity X of material Y, does that look right? And if I went back to first principles, how could I check that? If I want to say, okay, I’m not sure about some of the outputs I’m seeing here from this really clever model that’s doing this, I need to know from first principles, well, how would I measure that? As we all know, BIM models depending on how they’re fed, a BIM model can have elements in it that really you can’t build that. The BIM model may think they’re coherent, but actually you can’t build a concrete slab that big. So actually there’ll be that level of fundamental knowledge that we need to have so that when we’re using this kind of AI assistant, if I can use that phrase, so there’ll be times where we’ll say, okay, I think you might have got this wrong. Let’s go back to first principles and see actually, how would I measure that?

Paul Heming: No, I think that’s really, really interesting. And you know, sorry, QSs, I never like doing it being have to carry on taking off is what we’ve just heard there from, Andrew. 

Andrew Knight: Well, I think once again, there’ll be tools that will help you do that. But I think if you can’t manage what you don’t understand. So how can you manage a tool? For instance, we work with a tech partner who does lots of very clever use of AI in terms of scheduling and I think you’d still need to understand how to develop work breakdown structures and how you’d schedule a project rather than just being fed the output by a very clever machine. 

Paul Heming: Yeah. No, I completely agree with you and obviously being a little bit facetious there, but what I’d like to talk to you about is a phrase to just use there, actually the AI assistant. But we will do that right after this break.

So we’ve talked about the future of quantity surveying. We’ve talked about professional skepticism. You also shared, I like following the stuff that you shared. So I saw one of the posts that you shared, it was by someone else, but it was about a report by the Boston Consulting Group, BCG on AI. And there was a few takeaways. I think the chap actually had nine takeaways. I haven’t got all of those from, a few really kind of stood out to me. Right? So number one was that 80% of company leaders are currently using it, but only 60% of frontline employees are using it. Leaders are the most optimistic. 62% think it’s a good thing, 39% when it comes to frontline employees and more than one in three workers believe that their job will be eliminated by AI. Now, those kind of chime with what I would describe as my experience taking feedback on the podcast shows where we’ve talked about AI. I’ve had a lot of people, not angry, but kind of being very dismissive. I think it’s coming in some cases from a place of that worry of what actually AI is going to do. Now we’ve talked about the fact that we don’t think it’s going to wipe out quantities surveying, quite the opposite, right? I guess I’m interested to hear from you like your thoughts with regards to ChatGPT. I know that is very much simplifying AI, but it grounds the conversation most simply, I think, and I’ve heard stories about QSs trying to write scopes, trying to do X, Y, and Z on ChatGPT and seeing what comes out. I guess I’m interested to know what your thoughts as an organization and personally are on the impact of tools such as that on professional skepticism.

Andrew Knight: Well, I mean obviously the ChatGPT and the various other kind of generative AI products as they’re called have had a huge kind of public impact. And it’s hard to avoid them unless you’ve been on Mars recently. I guess for me it’s a kind of a funny sort of half impressed and half unimpressed in what they do because, and there’s a particular instance that happened recently, I think is quite instructive because these so-called large language models, by definition have been treated by, have been trained rather on a large number of documents. I mean, in the case of ChatGPT, a lot of stuff off the internet, which should as day one raise a few question marks on the basis that I’m not sure I trust everything I read on the internet, I don’t think I would necessarily use as somebody’s Wikipedia page as a first source of truth if I was going to use it for some kind of serious piece. And in the jargon ChatGPT can hallucinate it, it tries very hard to please when you ask it a specific question. And in fact, although it’s not QS related, there is an incident now in the states that’s well documented where a lawyer thought he or she was being heavily clever here by producing their closing statement or closing phrase, whatever it is, using ChatGPT. And it effectively hallucinated and produced fictitious case law, fictitious previous cases which he then bowled into court and tried to use. Yeah. And he was understandably caught out. And the judge, I think it’s fair to say, had a bit of a sense of humor failure now. 

Paul Heming: Really? 

Andrew Knight: Yeah. Oh yeah. I’m not making it up. I’m not hallucinating to use the phrase. And so this lawyer clearly got caught out by something as simple as not checking the sources and saying, well, are those real cases? And it’s back to that professional skepticism. These tools are not perfect. I mean, I’ve used ChatGPT a lot and I’ve used it almost day to make a point with people. I’ve done a number of speeches over the last few months on this topic. And I’ve literally used ChatGPT to write my presentations for me. I’ve been very upfront that I’ve used ChatGPT, and actually it’s done not a bad job, it’s got me 80% of the way, but I’ve checked it, I’ve said, well, actually, does that really make sense? No, I’ll take that bit out. So it doesn’t produce perfection, but it does save you a lot of time in the sense, I think all of us probably have that kind of blank sheet of paper moment where we’re thinking, well, I’ve got to write something, a report or a business plan or a, in this case a few speeches, where do I start? Well, I found ChatGPT to be very useful for that kind of starter for 10. Give me a few suggestions here and I’ll check them, I’ll make sure of the veracity of what you’re telling me and I’ll turn that into a finished piece of work. Hence that phrase assistant, they’re not the finished product. If you talk to some people who are behind these generative programs, they will admit they’re not entirely sure how they work. So that fact, and as ever, it’s the age old thing, punch cards, garbage in, garbage out, depending on what you train these models on, the result will depend on the quality of data that they’re trained on. And if we think of particularly the world of QS where you’re looking at cost estimation or benchmarking or analytics, then if you’re not feeding these kind of models with good quality data, that’s been almost curated by QS who say, well, actually, okay, well, this is the data we got at turnout. Yes, this is the correct data set. This makes sense. Yeah, I’m happy to load that into some putative AI model that we’re going to use. But once again, if these models are trained and/or fed with data that’s not correct, you will get garbage out. The cleverest algorithm in the world is not going to turn something that was incorrect into something that’s correct.

Paul Heming: Yeah, no, I mean, I hear exactly what you’re saying and that’s an interesting example, isn’t it, with the legal or poorly constructed closing argument that you were talking about. I mean, that makes perfect sense and that chimes with a lot of the things that you hear, it’s more like pub talk really as to what it could do for construction and surveying or whatever. I guess the question I have is as an organization, the RICS, how are you, it’s such a fast moving piece and I appreciate that it’s only really been a handful of months even now, where this has been seriously impacting day-to-day life. Like how, there’s a lot of talk at government level as to you’ve got to bring it into line, you’ve got to make sure that it’s structured in a way that it’s legislated for. As the RICS I’m not expecting it to be paralleled in any way, shape, or form, but is it conversation that you are having? Is there advice starting to be created around these kinds of tools?

Andrew Knight: Yeah, I mean, I think if I can return to the subject, I was talking in Toronto about the kind of automated valuation, which I know is not directly related to QS, but I think that’s at the leading edge of where we see not even generative AI, but machine learning, complex statistical models, producing output that our members might then be signing off and placing reliance on. And so I think for us, and this is perhaps to the earlier stage where we’re getting to think about what our members should be doing as a kind of almost a framework, a performance framework of understanding, well actually, if we’re going to get data sources into our work that come out of these kind of generative models or other statistical models, once again, it’s going to be an overused phrase in this chat. It’s this professional skepticism. Where did this data come from? Has it been, when I say manipulated, I don’t mean that in a pejorative sense, but has this data been analyzed by some kind of AI program? It’s come from some raw data, I’m now seeing some very impressive looking benchmarking numbers, but can I trace the data back? Where did this come from? Was it just scraped off a website? What kind of operations have taken place on this data to present what may look as a very compelling case in terms of the analytics? But that kind of continued, okay, where did the data come from? Has it just been pulled out of some, if it’s some major infrastructure project, have you just pulled it out of some government publication? Can we really trust this? Has it been normalized and rebased for the fact this data’s 10 years old and is Canadian dollars not GBP. It’s some basic stuff, but can we really be certain, not in terms of being a doubting Thomas, but saying actually no, can we just check that this data makes sense? Where did it come from? What kind of analytics have perhaps been applied to it? And then say, yeah, no, it’s great, but I’m happy to trust this because then critically, whether you’re on the contractor side or PQS, your advice is being relied on. So effectively you’ve got to stake your professional name and say, okay, no, I’m happy to sign this advice off at whatever stage of the project. And so for our own protection as a profession, we’ve got to have these guardrails that say, we should be able to ask these clever guys and girls who build all these tools and all these models. Look, I need to ask you a few questions, a bit like the end of Colombo. You know, I’ve just got one more question here. You know, where did you actually get the data from? 

Paul Heming: I used to like Colombo

Andrew Knight: That’s one last thing as he taps his head. But it’s that kind of thing of, well, because there is this illusion of accuracy that you get when you see something on a computer screen, oh, it’s come out with a number, $10 million for this. Oh, that must be right on a computer screen. 

Paul Heming: And that leans into your initial view on ChatGPT, where it is based. A lot of it is based on data from the internet. Do you trust all the data on the internet as your baseline? And therein is a little bit of professional skepticism from the outset, right? It’s really interesting and you’re talking a lot about data and your transparency around data there, what you’re talking about like where, what is the source of the data, et cetera, et cetera. I find that quite hard and quite abstract to picture how in the future you would be able to source that data and where it would come from. And then I’m also thinking about data in the context of what you’ve talked to me separately about with the ICMS and the fact that you yourself are head of data standards, you yourself are trying to champion the ICMS structure for data and really like bring that out to the market. Why should quantity surveyors know about the ICMS? Because I think it wouldn’t be known about in the same way SMM7 as a structure or NRM, sorry, would be right front and center. What is the ICMS and why do you want to champion it?

Andrew Knight: I think fundamentally, and I mean a number of reasons. I mean effectively it’s not there to replace NRM. I think there can be a misunderstanding that it’s there to replace NRM or SMM7 or CESM-4 or highways method of measurement or RMM. It’s there to sit above those standards because one of the challenges we have around benchmarking, particularly when you do that on an international basis, but even within country and jurisdiction, is how do you actually do an apples and apples comparison of costs, whatever stage of the project. Because If you try and benchmark when you’ve got a set of data in NRM against something in a different standard, clearly that’s hugely difficult because they don’t map directly together. So from day one, hence the name International, the aim was to put a relatively simple cost hierarchy on top of those very granular, very detailed costing systems, and make sure that we could benchmark that particularly for the larger the project, the easier it would be to get that sense of proper apples and apples comparison when we’re looking at getting best value. Because if we go back to kind of the ultimate aim of construction is to provide great assets for the public, whether those are government sponsors or private companies. We need to deliver value both in terms of functionality but also deliver on time and at cost and benchmarking, we all know that we could benefit from better benchmarking data and ultimately that’s what ICM is about. And it’s there to not try and get rid of these either local or asset specific type coding systems, but to sit on top of them with a common roll up. So I think what QSs should know about is that this exists as a reporting framework for them to roll up from all these different kinds of measurement codes and get a consistent way of benchmarking projects. And I think leading on from that, in its latter versions now with ICMS-three, it’s from an ICMS-two perspective, it was about getting whole life, not just looking at the construction phase, but predicting and looking at the maintenance, the refit, the end of life getting that full, whole life view of it. And then in ICMS-three, adding the carbon piece and really getting that full picture over the life of the asset. What’s the impact not only on cost, but the impact on both embodied and operational and when there is an impact from a kind of retrofit point of view? So that’s the important reason we need it. We need better data, better structured, able to be compared across border and across projects where we’ve used a different measurement system and to I think critically get the QS to think in that more whole life basis. And indeed to get project sponsors to think in that whole life basis and think, well actually, are there some tradeoffs here in terms of CapEx at the front end and design decisions that will give us better operational performance, better design life.

Paul Heming: Yeah. And to be honest with you, what you say about, it doesn’t replace NRM, SMM access and whatever. That actually really—

Andrew Knight: They’re designed to.

Paul Heming: Yeah. No, but it’s funny that you say people think that, that’s certainly what my first, when I saw it, not when I’d read it in detail, but when I saw it I thought, oh, another form of measurement or whatever, another standard of God really. But actually it makes total sense. We talked on Monday, didn’t we, about the fact that NRM might be your, if you think about it on micro – macro level, NRM, just sticking with that would be the micro, like the really detailed measure data sets specific to the exacts of the project. Whereas that then only filters into the ICMS and its coded back to the ICMS and that for you as an organization and for us as an industry in the UK and the wider international space is absolutely critical. Because we’ve got NRM, other countries have got whatever, whatever, whatever… And it’s not all feeding into the same data pool. So it’s really, really absolutely critical. And having now sadly taken the time to do quite a lot of research into it, honestly, or maybe not sadly, it makes a huge amount of sense to me that in terms of the quality of the data sets, and this will resonate with QSs, right? There are not many companies where you can say, right, how much does that cost as a trade or as a material element and you can go and get a 100% banker, I trust that data set and it’s exactly, you’ll have a mess from loads of different projects. And that kind of is effectively what you’re trying to filter into with the ICMS. So it makes a lot of sense. On the quality of historical data sets, talking BCIS, spawns, all of these different things, tools in inverted commas that QSS has had available to them. What is your view now, zooming out on the quality of the data sets that we have at our disposal as QSs?

Andrew Knight: I suppose perhaps rather controversially, I might say that they live or die by how the sector contributes to them. Because the weakness in any data set would be around lack of depth and breadth. If we don’t get enough projects into these kind of pooling services, if we don’t get the detailed breakdown of elements to give us that real granularity on projects, if we don’t get enough information about the attributes of the project to do a proper benchmarking. And it’s not that I’m putting the onus back on the sector, but what I am saying is that a lot of these rely on effectively information coming from the sector into a common pool. And I think it’s, perhaps if I can be so bold, a critique for the whole of the built environment that we’ve not been good at handing data in perhaps at least two dimensions. One of which is having it in a structured consistent form, but also perhaps not really getting the benefits of pooling data. But clearly every participant has a commercial imperative to make money to be profitable, all those very understandable things. But I think there’s been a kind of slightly empiric victory, a hollow victory by data not being shared enough. And I think particularly now, if we think of the context of sustainability and carbon and cost, I think there is an imperative for the sector to think, well actually there is a bigger picture here. Yes, we need to make, we need to be profitable to keep viable supply chains and viable contractors and tier ones and everybody in the whole ecosystem. But it’s really important now for the planet and for sustainability that we have as much benchmarking data, not just to produce our early cost advice and understand how to cost a project, but the impact of real estate means we need to have better quality data to understand the impacts of decisions we make, the materials we use, the embodied carbon, the long-term effect on design life, et cetera. And we have to find a way to share that data that doesn’t necessarily compromise commercial imperatives, but enables us for these data pools like BCIS to have a richer, broader, deeper pool in which to do their modeling to give better advice back to the sector.

Paul Heming: Wonderful. I couldn’t agree with you more to be honest with you. And therein lies the challenge, doesn’t it? Welcome to construction. There are some great organizations we sponsored on this show as well, project data analytics, community data trust, where they’re trying to break down those barriers, but those barriers remain, don’t they? Final question for you, Andrew. This is a broad and tough one. Where do you think quantity surveying will be in 2030 or in five years’ time? How will it be different? How will data and technology have shifted it?

Andrew Knight: I think the continued focus on sustainability, on it embodied and operational carbon, I think will have forced in the nice sense of the word, a lot more focus on data. For me, data is fundamental here because I mean, as a QS, that’s your lifeblood is the data that you work with. And I think there has to be that both this technical and this behavioral shift in the way we curate and handle and share data. So I would say because of this increased focus on sustainability, not only for new build, but for all this retrofit work that’s going to have to take place and also with hybrid working, the amount of change in use of floor plates now, the kind of amount of work that’s going to take place. I simply think that through that combination of the sector doing the right thing and arguably being forced to through legislation and various other kind of costs of the externalities, I do foresee hopefully this kind of democratization where increasingly down through the tiers of the construction industry, there will simply be more use of digital tools, more structured data, more tools that will actually help people and say, look, here’s a bunch of PDFs from this data, from this turnout of this project. Have a look at it, AI assistant and give me a summary of it. It probably won’t be 80, won’t be a hundred percent accurate, but give me 80% of it and I’ll tie it up and I’ll work on it and I’ll then be able to present that to my client or my boss. So I think that in combination with another generation coming up through the profession, we’ll simply get more comfortable with having an AI assistant. I mean, there was a chap on the BBC, I forget it was a tech firm not in the built environment, funnily enough, but he said, look, AI is not going to take your job, but it might, but somebody using AI might. 

Paul Heming: Yeah, it makes perfect sense, doesn’t it? And running away from it or pushing back on it is not going to safeguard the longevity of your career or your profession, right? So quite the opposite, I think you should put run towards it. 

Andrew Knight: I would see it as kind of excel on steroids, let’s use it as the next generation of tool that we now don’t see as a threat. We would be aghast if somebody took Excel off our desktop. I’d hope that by 2030 people would be aghast if somebody took their AI assistant off, which helps them with measurement, which helps them with benchmarking, which helps them with data analytics and things like Power BI that helped them present data in a nice simple format for clients. I think we want people to not want these toys to be taken away because they’re just part of their toolkit and they end up doing more exciting stuff than data set typing numbers into spreadsheets all day.

Paul Heming: Fantastic. And that is a super positive way to end the episode. Thank you so much for coming on the show and talking about your experience and clearly what’s something that is your passion because times are changing quite fast at the moment. There’s a lot to take on board. I think for quantity surveyors specifically I was very close to my heart and I think there’s so much to change and I think if you walk towards it as opposed to rally against it, you’re going to be in such a good place in a few years’ time. So what you’ve explained is only evidence that further, I will obviously leave your details in the description. You must all follow Andrew, because I enjoy very much the post that you put out there, really, really good. And I’ll share your details. Thank you for coming on the show and yeah, I’m sure we’ll speak again soon. 

Andrew Knight: Pleasure, Paul. 

Paul Heming: Thank you very much. And guys, as always, I will speak to you next week. Have a great weekend.

More Episodes