From Digital to AI: Hammad Chaudhry’s Vision for Next-Gen Construction Tech – Episode 11 | AEC Insights

00:30
Erin Khan
Hi everybody and welcome to another episode of the SureWorx podcast. Today I’m really excited to share that we are speaking with Hammad with EllisDon as well as Building Transformations, and we’re really thrilled to have him here in the studio with us to share a little bit about his experience with technology, innovation and all of the interesting projects and things that he’s been a part of lately. So, Hammad, I’ll turn it over to you to give an intro to yourself, your background and then also maybe if you can start us off with how you got into the industry, that’d be great. 

01:15
Hammad Chaudhry
Thanks, Erin. Yeah, absolutely. So I’m working with EllisDon as the Vice President of Innovation and Construction technology. So really my focus is on working with a few of our different teams. So BIM, VDC, Construction Technology and then some of the innovation efforts with startups and the accelerator programme that we’ve launched. And my background to get into the industry was originally I wanted to become an architect. So I started working in architecture and being based out of Alberta. There was a little bit of a recession that came from the oil and gas economy around 2012, 2013 or so. But construction was really doing well at that point. 

01:55
Hammad Chaudhry
So I decided to, you know, take my friend’s advice who was working at EllisDon, he got my resume there and then I got interviewed and next thing you know, I joined EllisDon as a project coordinator and a BIM coordinator and it’s been almost 11 years since. 

02:11
Erin Khan
Wow, amazing. I’m curious, why the initial interest in architecture? 

02:20
Hammad Chaudhry
Yeah, I don’t know, I guess it’s. It’s a good question that I don’t have an answer to. Yeah, I liked computers, technology, design, you know, cars, buildings. So I guess it’s just something that is attracting me and then the aspect of architecture that, you know, as I was growing up really kind of made me realise that there’s an opportunity there was when you travel to different places and you see kind of vernacular architecture and then you see kind of some of the more modern architecture and how that changes how you feel. Right. So if any North Americans have been to Europe or Asia, that immediate feeling that you get and how architecture can impact that’s maybe what got me there. But it could be just a little bit of fluff, you know, looking at the rear view mirror and putting pieces together. 

03:09
Erin Khan
Yeah. Well, I’m glad we have you on the construction side, at least of the industry. So. Yeah. So tell us a little bit more about, like, EllisDon. So you mentioned an accelerator programme. 

03:23
Hammad Chaudhry
Yeah. So EllisDon is actually probably one of the coolest Canadian business stories, in my opinion. So construction company started by two brothers out of London, Ontario, which, you know, very early in their kind of journey as a company, they bet their future existence on tower cranes. So they were the first construction company. Yeah. In Canada to purchase and bring tower cranes in. And so, you know, the kind of entrepreneurial spirit, if you will, has been baked in, I think, from the beginning. And then also for me, kind of looking over EllisDon’s kind of past projects and what they’ve done to build Canada is also hard to ignore. 

04:08
Hammad Chaudhry
So some of the kind of monumental structures like the Sky Dome or Roger Centre, which is where the Blue Jays play to a bunch of the healthcare facilities, you know, the public private partnerships in Canada, they really were one of the major players to drive that cross. So that’s kind of the background of EllisDon and kind of some of the history. And the accelerator programme started about two years ago now with a little bit of a different vision and angle to say one of the things that EllisDon has that maybe not a lot of companies have the opportunity is the footprint and the variety of geographies that we work in coast to coast across Canada. You know, if you’re working in British Columbia, say Vancouver, it could be starkly different than Alberta or Ontario Maritime, so on and so forth. 

04:55
Hammad Chaudhry
So what we wanted to offer at the accelerator to startups is you can probably find product market fit a lot quicker working with us, given that broad market kind of purview that Alliston has. And on our side, then we’d be able to kind of work with some of these startup providers at a really early stage as they’re developing and maturing. So kind of a win proposition, in my opinion. 

05:20
Erin Khan
So there’s a lot to unpack there. One of the things I’m interested in is maybe exploring, so I think one of the things that’s unique about the perspectives that we have here, like, you’re in Canada, Nandan, you’re in Australia. Some of the parallels is, again, the geographies are, you know, very, there are some that are very harsh or difficult to build in, or a wide variety. Right. Could you speak a little bit more about how that influences innovation or throughout company? 

05:56
Hammad Chaudhry
Yeah, that’s a good point. I guess part of it is going to be, hopefully it drives us to be more pragmatic, meaning there is a lot of technologies that some of our peers may have deployed or done a little bit of innovation theatre, if you will. And although we could do some of those things, I think, to your point, given the geography and some of the constraints that we face, it does make innovation perhaps a little bit more practical. One of the things across Canada that’s kind of a joke in certain kinds of political circles or business circles, is Canada likes the US or the UK to try things first and then once it’s been proven out, we’ll do it our way. 

06:39
Erin Khan
Right, I see. 

06:41
Hammad Chaudhry
Yeah. 

06:43
Erin Khan
I haven’t heard that one before. So thanks. Thanks for sharing that one. 

06:47
Hammad Chaudhry
We got to get both of you into Canada more often. 

06:50
Erin Khan
Yeah, yeah, for sure. We’ll come visit sometime. 

06:55
Hammad Chaudhry
So I think that does have an impact. And also to your point, because of some of the constraints, it may change the focus on the technology as well. So a lot of the bin VDC kind of workflows in some of those more remote areas make a lot more sense. So you’ll see industries like oil, gas, process piping that are significantly more advanced in the way that they integrate those technologies. But that doesn’t mean all northern communities. We still have a lot of differences, if you will, between some of our more northern communities and the way that their trades and the workforce integrate these tools and technologies compared to our urban centres like Calgary, Toronto, Vancouver, for example. So I’m not sure if that fully covers it, but there is definitely an impact, I think, from the geography. 

07:44
Erin Khan
Yeah, yeah. 

07:45
Nandan Thakar
I mean, just drawing parallels from my side, Erin, you mentioned Australia and Canada. 

07:52
Erin Khan
Yep. 

07:53
Nandan Thakar
A lot of similarities between the two countries. I mean, landmass is one, but population dispersion, how harsh the conditions can be through the calendar cycle. And yeah, Hammad’s point, I second that, because being a CIO of a company of thousand people with remote projects, the first gate that a solution needs to cross is does it work in offline mode? Which is not a question you ask of a solution built for metro cities and high rises. So not only do you have remote locations where just logistics of getting people and planned there is one item, but then there is the connectivity and life is a bit better now with Starlink and such. But being able to work in offline mode is one. But then also linear projects, right. 

08:53
Nandan Thakar
So I think the use of technology where the location like a hospital is fixed is one thing, it has its own challenges. But try using tech for a pipeline build through harsh conditions for three kilometres a day and the tech that is used in a linear project where the location and the camp moves every day is very different. So we see a lot of cars, you know as an example. So if Tesla or BMW have a new car in the pipeline they would send it to Australia to do their 50,000 miles with the stickers on. So you know it’s incognito mode. But they just test the car out here before they go. It’s ready for a European or North America market. So yeah, it does allow especially Australia to be that POC centre for new solutions. Yeah, so yeah. 

09:57
Nandan Thakar
So I think from my perspective Hammad, I would love to get a bit more feel for our listeners as well in terms of kind of project that your company embarks on and then we’ll get into the challenges and the solutions that you are trying to explore. 

10:16
Hammad Chaudhry
Yeah, absolutely. So EllisDon right now is building a significant amount of hospital projects across Canada. These are some monumental jobs. So you know I’m talking some of them on their own are north of a billion dollars in project value. Some of them also are so large that we’ve joint ventured with you know, PCL for example on the Trillium project out in Toronto. So healthcare is a massive focus right now. Just for some of the folks if they’re listening. EllisDon’s working on Cowichan Hospital in Vancouver Island, East Surrey Hospital in Newton, Surrey, British Columbia. Royal Inland’s being wrapped up. Burnaby Hospital, we have South Niagara Hospital, we have another one in Grimsby, Ontario, West Lincoln Memorial and then there’s a couple of other projects in Ottawa as well. So there’s the CHO project and so these are just a few that I remember. 

11:23
Hammad Chaudhry
Obviously it’s like off the top of my head but these are monster jobs and so that’s a big focus. But we also have a big amount of condos and kind of mixed use residential towers as well as you know there’s that big boom in Canada over the past several years. So in Vancouver, in Calgary and even Ottawa, Toronto those have been a pretty big driver. And then you know we have a Life sciences division, which is focused on the pharmaceutical, the data centres. And that’s another segment that of course has grown significantly. One thing that EllisDon typically, I’ve noticed, is able to quite, you know, meticulously serve in the market is they can design, build, operate, maintain, finance, equip, almost any type of project. 

12:15
Hammad Chaudhry
So, you know, if it’s a civil job or a renovation or one of these hospitals we’re talking about, that’s kind of all within the wheelhouse. 

12:24
Nandan Thakar
Very good. And yeah, that sort of brings us to the scope of challenges you deal with because you seem to have multiple verticals of what you service, from civil projects through to refurb projects and residential and commercials. So you’re sort of doing, you know, all those four things within the spectrum. Yeah, just give us a feel of kind of solutions that you’re looking for out there in terms of challenges you face and what are you most excited about in terms of where the solution landscape is going. 

13:07
Hammad Chaudhry
So some of the things that, you know, we’ve been focused on for the past maybe two or three years right now has been around some net new technologies. So startups that are providing tools or methods that were unable to do before, as well as kind of some core tools or processes that may not be as, you know, shiny, so to speak, when it comes to all the hype behind AI, robotics and all that, but are very practical, pragmatic. And so I’ll start with that side first. So we have launched what we call kind of our construction tech ecosystem, which is looking to complement the stuff that EllisDon builds in house for software like our project management system with some of these other tools and processes. 

13:57
Hammad Chaudhry
And when we look at one area, which I think is quite obvious, but obviously difficult to kind of implement and whatnot is going to be around scheduling and planning. And so that is something that sometimes you’ll see people, you know, marking up whiteboards and trying to coordinate other tent stickies. Some people like Lean and are bought into that methodology and trying to make that continuous improvement. So something that can attack that problem of helping people not only communicate but collaborate and effectively keep track of the schedule and what’s moving. So we’ve partnered with Touch Plan over the past few months as one of our strategic technology partners there. Very similar to that scheduling and planning would be logistics and deliveries. 

14:47
Hammad Chaudhry
So if you think about one of the hospitals that were talking about, I’d say if it’s a two year job, we could see 14,000, maybe more deliveries coming over the course of that job. Right. Maybe a little bit more. And typically same thing, it’d be through phone, email, maybe some sort of shared calendar. So we partnered with Structsub and Structure is focused on digitising that process. You can have a mobile app, a web portal, so on, so forth. So those are kind of some more pragmatic examples of tools that I think are going to be effective and help move the needle, but may not be as cool and flashy as some of the AI things and whatnot. On the flashy stuff and the AI side, because I know that’s kind of interesting to people as well. Right. Is going to be. 

15:37
Erin Khan
We were definitely going to get to it if you didn’t bring up, here we are. 

15:43
Hammad Chaudhry
So there you go. We have. I mean, I could go back. There’s this one company we work with in 2019 and they’ve been doing really phenomenal work in the market ever since as well. Is AI clearing the premise there? The two gentlemen spun out of, I think, PwC’s drone division and started creating a platform that could take drone data compared to your schedule, your civil 3D files and start to really give you pulse updates in a way that we hadn’t seen before on your civil infrastructure jobs. And the idea there was if we are flying drones and we’re doing cad work, civil 3D, all that kind of stuff, and we’re already, we have a schedule and really project controls is the key when it comes to a lot of these kind of jobs. 

16:30
Hammad Chaudhry
With something like AI clearing that, you know, was talking about AI and whatnot before is kind of cool, be a practical tool or not. So we tested that out concurrently. So we’re running AI clearing, collecting data, getting the reports, but the team is managing the process traditionally as well. And so were able to baseline and compare how that kind of went and over that kind of proof of concept and pilot, I think we did that for nine months or so. We were able to say, okay, there is a big opportunity actually from a tool like AI clearing the marketplace. 

17:05
Hammad Chaudhry
But the biggest downfall for Canada is we don’t actually have that digital information as a need or a requirement from the design firms traditionally for a lot of these infrastructure projects that the government is building and that’s where the most value would come, right? Like kilometres of highway, road, utilities, so on and so forth. So we had to do all that front end work with our team, reproduce all that content to be able to leverage a system like AI clearings. So culturally we have a lot of work to do. I think in Canada, but maybe even North America to even leverage technology like AI clearance. So that’s a civil example. We could get into others if you want. 

17:45
Erin Khan
Yeah, yeah, that, that’s a lot. Great, great examples. I’m. I’m curious. So, you know, we’re. You’ve brought up something that’s pretty like futuristic, right? Or I guess it’s now AI capabilities versus, you know, something that just helps you digitise your schedule or, you know, do lean pump planning. That’s a little bit more practical application. How do you guys balance, or. Yeah, I guess, let me see how I want to ask it. 

18:18
Erin Khan
But what I’m noticing through a lot of the conversations, and this one as well, is that there’s kind of a tightrope between what’s actually useful today to my worker in the field, who just needs to get a lot of stuff done to get the project built, versus what do we need to anticipate coming down the pipeline that we need to start experimenting with so that when it actually does get to the point of being, having that, you know, pragmatic value, we’re ready to capitalise on that opportunity. How do you balance the future with the now? 

18:58
Hammad Chaudhry
I don’t think there is a one size fits all part of it. You know, I don’t want to get too theoretical or, you know, business kind of NDA kind of thought philosophy, but I would argue it comes down maybe to the company and their culture. And if that culture of the organisation is, look, we need to be forward thinking, we need to be innovative, we need to be delivering solutions to our clients that take risk away or stay up with the market. There’s going to be a need for you to test things out and fail. 

19:33
Hammad Chaudhry
And so I think really the balance would come out to be able to validate some of those things that we want to have, say in five years or 10 years, and have the expertise and make sure we’re market leaders in a way that doesn’t hinder current performance or financial success. So, in other words, if there was a job where maybe we could do something that wouldn’t be typical right now, but it is maybe more futuristic, like, say, robots on site or whatnot. And if that job doesn’t have the right risk profile and whatnot, then that’s not the right place for us to be doing that. 

20:12
Hammad Chaudhry
Other times, you know, there’s creative avenues as well where you can work depending on what this kind of initiative is that we’re talking about with the different business leaders, so that you actually get support a little bit more from the folks who are looking at the business plan, their performance and some of those things. So now you’re not trying to balance that on your side. You’ve actually got support from people who know what their book of business is looking like and where there are opportunities for us to maybe test some of these things a little bit more effectively. And then the last thing, Erin, I think that works pretty well is there’s opportunities with the government. So we have like National Research Council in Canada. Another thing, like my tax is like a grad programme, so you can also explore those avenues sometimes. 

21:01
Erin Khan
Interesting. Yeah. The government resources or programming is not one that comes up often in conversation, at least in mine. So thanks for expanding that viewpoint for me there. 

21:15
Hammad Chaudhry
Yeah, no worries. I don’t know if this is a bad topic, but I think it might be also a difference between Canadians and Americans as well. In terms of government mandates. 

21:26
Erin Khan
Perhaps. Perhaps. Well, yeah, there are some differences for sure, but I guess strengths and weaknesses on both. 

21:35
Hammad Chaudhry
Yeah, yeah. 

21:36
Nandan Thakar
Funny enough, we might have more similarity 15,000 miles away with Commonwealth Partner Australia. But just on that topic, I was on an innovation panel on the topic of innovation last week in Sydney and we explored that topic essentially. And where we, as SureWorx as a vendor have landed is that if we have firstly, we are a platform. So you know, we’ve got lots of different modules that bring in data and the platform strategy for us, being an ex CIO from construction industry made sense to me rather than a whole lot of point solutions. Best of breed, which is fine, right? Sometimes best of breed is fine and sometimes a platform solution is much better, especially if it’s modular. 

22:27
Nandan Thakar
But once you have either an integrated best of breed toolset or a platform, the opportunity that we see, and I’ll digress a little bit, I don’t know if you guys saw the org chart that Nvidia published in their update last month, but they had a human org chart with AI agents reporting into people. So AI agents and humans coexist within the org structure. And then if you raise eyebrows and then once you think through that and you go so just thinking through that, right? Like in construction we are used to people and plant and underpinning processes. And then you have robotics which is starting to become repeatable. You know that there are some functions that can be done if your site is ready for with robotics and what we are doing with reality capture and all that. 

23:30
Nandan Thakar
But then you overlay AI agents and this concept of co-pilot. But the way we are thinking about this is, let’s say we have half a dozen modules from quality and construction commissioning to health and safety asset lifecycle management. So, so if were to aim for AI agents across each of those capabilities, then those agents are like a portfolio that is available to you, like an app store and then you can trigger them once certain conditions are met. And so if I am a site superintendent, then I will have three or four agents that I can opt in because they are available in the menu that will help ease X number of things per day, per week for me. 

24:16
Nandan Thakar
And so that was not that Nvidia said it, but we have been at this for six, eight months and I’ll give you a few examples. Right, so in quality and commissioning we have an AI agent that summarises your schedule and then it sends you an email to say these are the quality cheques which are due today by X number of subcontractors. And then if there were any slippages in schedule or non conformances that sort of gives you a summarised output. And we have seen that to be of value that you are automating smaller tasks which make minor decisions based on some training that help everyone. So yeah, just an interesting thought process where you have people plant, you have robotics and then you have AI agents and that is almost like your construction workforce of the future. 

25:10
Nandan Thakar
Any thoughts or is that where you see that going? 

25:16
Hammad Chaudhry
Yeah, I mean there’s a lot there and I think I’m not even going to try and say I know all of the things that are happening when it comes to your point. The agents and some of the stuff that they’re doing. I was at a conversation the other day and they’re talking about Nvidia and how Nvidia’s kind of got these new GPUs that have a bunch of these open source models that are really refined and more effective to run on their GPUs and you can bring them in, build your own server rack and have all these LLMs. And then they talked about this concept of rank or retrieval, augmented generation or whatever to help stop the hallucinations and so on and so forth. And then from there to your point, what you’re talking about agents being able to deploy these agents. 

26:02
Hammad Chaudhry
So in theory, I mean, I guess the concept makes a lot of sense practically from a deployment standpoint. I think the number one thing that we’re going to have a little bit of confusion or resistances, the black box kind of feeling. So already when we’re using simple tools that are not really AI in the sense of anybody in deep tech would consider AI, but maybe just an average person who’s not in tech would think it’s revolutionary. Say, you know, the tools like stress hub or whatever, or even tools like cubics and buildots and whatnot. People already feel like that’s a black box that they don’t understand and is there any decisions going on that they don’t know or were missing? 

26:45
Hammad Chaudhry
And so when you bring in these agents now and we talk about a superintendent, sometimes, you know, that trust factor is going to be really hard to overcome. And I think that piece is something I don’t know how we talk about. For the technology side, I think there’s a lot of smart people in the world working on debugging, fixing Iterator, but for that adoption side, I think that’s going to be tough. Right. How do we get over the black box approach? What do you think? 

27:13
Nandan Thakar
Totally, yeah. Look, I’ll give you an example on your point and that is very true that how do you bring trust and how to make sure that decisions are not driven without oversight? And an example that were exploring earlier, which is in fact an opt in capability we have, is that in health and safety, typically you take a lot of photos as part of any incident reporting or observations or whatever processes you have and you could end up with hundreds if not thousands of photos per site. And if were to train one of these AI agents as a use case on identifying hazards in that photos, but they don’t make a decision, they present that decision to the health and safety officer to say, we see this electrical hazard, do you agree? 

28:15
Nandan Thakar
And so that way there is a system where you are training that model and we trialled this with a customer over six months and they got over that trust factor to say, I think this is actually helping me because what AI is good at is needle in the haystack stuff. And if out of 5,000 photos they picked up two hazards that the human eye missed, then that is good tagging. And then the next step is, well, if we do identify a hazard that you validated, would you like us to log a ticket for somebody who is on that site? And so that is minor level of decision driving, but it is better than your robotic process automation, which is essentially a bouncing ball. So the fact that you can identify 10 containerized hazards is seen as a net benefit. 

29:07
Nandan Thakar
But it took six months to get there and now we can industrialise. So perhaps that, you know, just giving you a practical example of theory is good. But bringing people along baby steps, earning their trust, making sure that the model is being trained by experience and people who’ve been there, done that. And then they harness that as a report to say, well, which is my safest site? And then it’s based on all this, it says, well, out of 11 sites, site two is your safest for this reason, that benefit. 

29:42
Hammad Chaudhry
Yeah, a very good example and very practical. But one thing I’ve got to ask you. So a lot of you probably heard this as well, Erin, people say, you know, this AI stuff came out and it’s really cool. ChatGPT, Cloud, Llama, whatever. 

29:57
Erin Khan
Yep. 

29:57
Hammad Chaudhry
And now I can write an email or a poem or do art, create all this cool stuff, but I kind of enjoy doing some of those things. But it can’t do my dishes, for example, or it can’t, you know, do some of these other tasks. So same thing within construction. I guess my question to both of you is how are you seeing the deciphering of tasks that are valuable for humans to still take on, given the risk or maybe the relationship or just being able to manage the chaos as a human more effectively versus which ones really do make sense for AI to take over and finding that balance as well? 

30:34
Hammad Chaudhry
Because I think the trust piece you mentioned makes a lot of sense, but that value prop piece too, I’m curious, where do we want to retain control as people versus maybe trust these systems more? 

30:46
Erin Khan
Yeah, I guess at least a thought on that is especially for the human versus AI, in which tasks do you feed one versus the other? I first start, want to start with just a scale of data. So there are certain things that a human brain will never, ever have the capability to be able to process. So if we say to like you, Hammad, okay, you know, sit down and let us know everything that’s going on in terms of change orders with all of the hospital projects that are happening across all of EllisDon, with every different region. And there’s no way that, you know, in one instance a brain can just, a human brain can just process that. But for an AI, it’s designed to process massive amounts of data and be able to, you know, quickly pinpoint certain patterns and trends and insights. 

31:53
Erin Khan
So I think scale is going to be one of those things that helps indicate, you know, whether or not it’s, you know, it’s a good litmus test. Like, is this a good exercise for a person to do or is it a good exercise for my AI agent or, you know, whatever tool it is. So I think scale of the problem or you know what the data set is going to be indicative of some of these exercises. There’s probably a few more that if given a little bit more time I could probably put some framework together. 

32:29
Nandan Thakar
But that, yeah, look, I’ll add to that. I think scale is certainly true. You need volume of data and AI, if done right, seems to excel at unstructured data and then suggesting a decision, not making a decision. So I think that’s where we are. But also complex scenarios, right? Like so if I were to digress a little bit, they are finding pharmaceutical industry is finding drug combinations much faster because of the permutation combinations you can go through with AI in cancer research you are finding the patterns better than somebody with a white coat in a lab for 10 years. And there’s nothing against that human effort, but there are just some things that AI could do better at the pace at which it can go. So I mean that’s obviously the promise. 

33:30
Nandan Thakar
But you know, we gotta be very conscious of data not leaking out, that there is no inbuilt bias into the AI that continues from the human team that built it. So yeah, it’s a double edged swot. But like with most things, if there is right parameters around it and it’s done with consideration for people on the floor and it’s like an adaptive thing to say, well this is what we can do, what do you think? Rather than we must do this, then it could be a net benefit to I guess having been in construction after being in it for 20 years. The thing that staggered me when I joined construction in 2018 was how tight are the margins in this business, right? You go, why do someone do construction? You wear all the risks. 

34:28
Nandan Thakar
The margins are, you know, it may not touch double digits if you’re lucky and if one risk eventuate, there goes your margin. So why do it? And obviously the answer is, you know, it’s a nation building infrastructure, someone’s got to do it. But how to improve productivity and margin could be a byproduct of it. How to just improve productivity, how to improve safety and how to improve work life balance where you can actually clock off at 3pm or 5pm once your shift is done and still achieve what’s there in the schedule. So to me, if there’s any contribution this trend can make to that safety margin and risk improvement, risk management improvement, it’s incumbent on us to at least explore it. 

35:18
Hammad Chaudhry
Yeah, yeah, there’s a lot, a lot of good points there. But I guess another question for you then. If you don’t mind? And I know this is kind of reversing the tables here, but. 

35:28
Erin Khan
No, this is a great discussion. 

35:31
Hammad Chaudhry
Yeah. In the reverse, what about like the piece of ethics? Because obviously that’s a big consideration and maybe a little bit more apparent when it comes to maybe government use or some of those things that have like, you know, a little bit more public eye. But even within our industry, whether it’s architecture, engineering, construction, what are kind of the best practices or is there any conversation around that? We were at a conversation a couple months ago in Chicago where Bluebeam CEO Smile Shuja mentioned how their approach is to be regarded as the organisation that looks at AI through the ethical lens as opposed to speed or first to market necessarily, because there’s a lot at stake potentially if you are first to market and you do not have those considerations figured out. So how is that within AEC being looked at? 

36:26
Hammad Chaudhry
And what are some of those metrics, for example, on your side that you’re building these products? Or even Erin, with you guys who are kind of consulting in the market, how. How do you look at that? How do you see that? What are your thoughts? 

36:39
Erin Khan
That’s a really big question. I think a lot of organisations and individuals are asking themselves as well and kind of coming up with, I’m not sure of the answer. I think what I’ve seen at best, and this is a really good first step in the right direction, is that some organisations are putting together governance boards around AI, specifically within their organisations to help have some measured policy, to be able to say, okay, this is responsible use of AI, this is not responsible use of AI and guide team members to down the responsible use of the tool. So I think that’s a big first step and an important first step that a lot of organisations are taking in the right direction. 

37:32
Erin Khan
I think one of the pieces that’s missing is really some sort of entity or group of experts that’s really letting us know, here are the actual ways that it’s working on the backend. Here are some of the practical steps that you need to do in order to use this tool ethically. Here’s what you need to know to be able to weed out the unethical models versus ethical. I don’t think we have answers to those questions quite yet for our industry and I’m not even sure if we’re asking that enough to be able to get those answers. So I like that you’re bringing this up because maybe we all need to ask each other a little bit more often before we dive into our solutions. Like what are the implications of this solution beyond just the application of it? 

38:27
Nandan Thakar
Yeah, yeah, so, and that point around ethics is central, but ethics is a spectrum, right? It’s not black or white. And I think if we rely on our good friend Elon, who have solved a lot of these things, his experimentation with FSD is where robotics AI come in. How do you replicate human driving? And I think we’ve all seen a lot of those videos. I saw one recently where there’s a very narrow path in a very touristy place and two cars are going on both sides and a 60 year old looked like an older gentleman just fell off, had a heart attack or somebody straightaway fell onto the road and cars were travelling at good 20 miles. Right. 

39:19
Nandan Thakar
So and this FSD driven Tesla made a decision that it needs to save human life and it turned at an angle and crashed into an oncoming car. So yes, two cars were damaged but that person wasn’t run over. And so you know, it’s a situation where it’s like, well I’m faced with two scenarios. One is a car crash where you know, if it done at 20 speed the, you know, the safety system will kick in but the person is safe and that’s exactly what it did. So the way he solved it was just to take all the lidar and cameras and everything out. And so there was a reboot on the FSD programme 18 months back where he’s like, we are just going to replicate any Uber driver that has five star rating, right? 

40:16
Nandan Thakar
So they just followed a five star rated Uber driver on the decisions they’re making in different conditions and light conditions and people and you know, someone cross and then they just replicated that because a good driver, a good human driver times 10,000 of them is the training that the model needs. So where I’m going with this, I think, you know, a good driver is an ethical driver, a good driver is a law abiding driver, right? Like there are some common traits of what makes a good driver and same goes with a good construction person on site that what are the decisions they are making in their surroundings that we can train a model on? 

41:03
Nandan Thakar
On just going back to that identification of health hazards that you know what makes a good health and safety person that well, you know that they have an eagle eye on any risk with hazard and if you are training the model through their encouragement and approval or review then the model is going to get closer to what good if not better looks like. So perhaps that is the answer. But again, it is one size doesn’t fit all and it’s a discussion that will, I’m sure happen over a long period of time. 

41:41
Hammad Chaudhry
Awesome. 

41:41
Erin Khan
Yeah, this has been really good. I wanted to share one of my takeaways actually from a little bit earlier in the AI discussion about trust and it just kind of summed it up pretty well. So I feel like the execution of our AI applications for whatever task we’re going to be doing or looking at in construction or otherwise, it will be limited as far as the least trusting person, I guess, in the group. So it all comes back to our confidence in the tool. So it’s only going to be as applied or well applied as much as we trust it. Right. 

42:31
Hammad Chaudhry
So that’s a, that’s a good summary. It kind of reminds me of even BIM and BDC, even though it’s been around for 10, 15, 20 years depending on who you talk to. 

42:45
Erin Khan
Yeah. 

42:45
Hammad Chaudhry
Still a lot of the use isn’t there because that trust isn’t there for them. From the boots on the ground or that soup or PM or whatnot. Right. This is an interesting parallel that even though it’s a completely different technology, culturally we’re seeing the same kind of mechanism that you’re talking about, which is trust and trying to build that up. 

43:09
Nandan Thakar
One of the things that. Sorry, Erin. 

43:11
Erin Khan
Oh, no, go ahead. 

43:12
Nandan Thakar
Yeah, I just wanted to just take that same stream of thought and more at an organisational level that we explored in our other discussion too. Hammad is. We’ve discussed a whole lot on the potential of AI and other solutions and how that may help with risk management or whatever. I would love to know your thoughts on somebody who is at the pointy end of identifying the solutions and doing POCs and making them viable in terms of business value. What are your observations on how a promising POC can translate into a company wide solution in terms of organisational change management? Like what are your observations so that a good idea actually settles in, that becomes our ways of working, like iterates the DNA the right way. 

44:11
Hammad Chaudhry
This is a very good question. That’s really hard to give you a straight answer because I think there’s so much variety. So every single company I’ve talked to, even not in construction. Right. So some of my friends are in finance or manufacturing and whatnot. The number one pain point I’ve ever heard is ERP implementations are never easy, they never work. And when you think about it’s really weird because that’s like core to any business, right? You think about it as like a core function, it becomes the backbone. And yet that is somewhere we have so much friction. And it’s known there’s probably in subreddits or memes and stuff like that, probably about European implementations at this point. But then on the flip side, you have other tools that maybe almost grow like wildfire, right? 

45:03
Hammad Chaudhry
Like you know, you’re trying to go through control process, this is how we should, blah, blah, plan it out. And it’s almost like it ignores all those. The people love it so much that just takes off. And so I feel like there’s like, you know, that spectrum of really you can plan it, you can be, you know, you can have method, you can train people and all that. And still it’s going to be really hard because there’s just so much friction for whatever reason. And I don’t have the answer there. On the flip side, some of the tools, you know, almost make it easy for adoption. The friction is on the value and the users just want to use it. 

45:43
Hammad Chaudhry
And so I think one of the key things is trying to understand, and this is hard to do is what kind of a solution, as best as you can gather, is it that you’re trying to introduce now or scale up or whatnot, is it going to be more towards that difficult friction kind of fill side of things or is it kind of showing you it’s a little bit more of that organic and people are really wanting it? I think that is very simple. It sounds but can change your approach differently because it comes down now to how you interact with the people, the stakeholders, the end users is going to be completely different on both ends of the spectrum. Then the other thing too is a lot of times this sounds maybe counterintuitive, but it’s not really me or my team. 

46:31
Hammad Chaudhry
It is actually the business itself, the people who build, who operate, who know the business very intimately. We’re the ones that are guiding those decisions and where they’re supporting putting guardrails in place, making sure things are kind of clear. But really when that happens, when you get that kind of buy in and traction, it really changes the outcome, right? So that’s another thing too is you can force it if you want or you can get a tool that’s supposed to be really good, but it’s full of friction. And, and no matter what, I think it’s going to be tough, but you’ll get there because you have to as a business. 

47:08
Hammad Chaudhry
On the flip side, you can have tools that really impact people quickly, no friction, you have support and it becomes their idea, essentially, and they want to champion it further and help get that implementation. So it’s not a holistic answer of good examples, but gives you the spectrum. And if you’re in that, I think, situation, hopefully you can use that guide to help figure it out and maybe give Erin a call if you need some more proper guidance. 

47:38
Erin Khan
Appreciate the setup. 

47:40
Nandan Thakar
Yeah, that’s, that’s as good answer can get, to be honest, because an organisational change management strategy with the right comms and committees, if, if the business unit is not equally keen, then you can’t force it down their throat. And then your other point is quite profound, that, yeah, you can have an org change management and you should, but if the tool makes it easy for it to go viral and adopt and easy to absorb and get going, well, that helps too. So, yeah, very good. I like it. 

48:19
Erin Khan
All right, let’s see. Hammad, anything pro-tips for our budding innovators out there that you might want to share. 

48:29
Hammad Chaudhry
Yeah, I think one of the most helpful things, and I mean, I think this is probably universal, even if you’re not in an innovation role, is when you’re in your organisation, you may not have too many people that will understand what you’re trying to do. Right. You’re in a little bit of a unique role. So how can I talk to somebody and try and get maybe some perspective, insights, help, even maybe exchange notes or for the lack of a better word, just spitball. And in my opinion, the best way to do that is the industry. So, you know, as you mentioned earlier, I’m part of this organisation building Transformations, which is a not for profit, led by industry. All the boards volunteer and really the only MO there is to move the industry forward. And so through things like that, I think would be. 

49:26
Hammad Chaudhry
My advice is really, there’s a whole tribe of people out there, you probably just haven’t met them yet, but if you try, you can connect. And that’s probably been one of my biggest kind of helps, to be honest, is I just talk to these people and get your advice and learn from them and a lot of people are willing to share if you’re actually genuinely interested and you ask the right 100%. 

49:53
Erin Khan
The industry is simultaneously so big, but also quite small. So, yes, once you get to know the, I guess, people who are interested in the same things you are, it can be very centred. There’s a lot of synergy. 

50:08
Hammad Chaudhry
Yeah, yeah, absolutely. So, yeah, that’s what I think. Yeah. 

50:13
Nandan Thakar
Thank you. And that’s the intention with our podcast as well, Hammad, that we bring those discussions to light where they can branch out into considerations at corporate level or forum, community level, in whatever format it takes. But it’s just sort of voice of the customer, the voice of the general contractor and the voice of, I suppose, the solution builder in terms of a software or robotics vendor, then, because it’s that trifecta that has to find efficiency, that create your destruction and land something that is valuable and sticky. 

50:57
Hammad Chaudhry
Absolutely. 

50:59
Erin Khan
All right, so I think we’re getting close to our time here. Any last thoughts, Hammad, Nandan for today’s episode? 

51:09
Hammad Chaudhry
No, all good. This is great for me. 

51:11
Nandan Thakar
Thank you. Yeah, likewise. Lots to ponder. It’s been a great discussion. 

51:16
Erin Khan
That was such a great conversation with Hammad. What really stood out to me were these key takeaways I think you’ll find valuable. First, partnerships drive innovation. EllisDon’s collaborations with startups are a game changer, helping both sides grow and innovate faster. Second, regional challenges demand unique solutions. From remote locations to urban centres, tailoring technology makes all the difference. Third, it’s about finding balance. Practicality meets cutting edge tools like StructSub simplify workflows while advanced AI opens doors to new possibilities. And finally, trust and ethics in AI are everything. Building confidence through transparency and responsibility is key to adoption. These insights are so inspiring and I can’t wait to hear how you’re applying them to your own projects. 

52:06
Erin Khan
Thank you, Hammad, for joining us today. Really appreciate your time and your insights and the discussion. I’m going to have to ask myself some questions about ethics later on, so I’ll be pondering that question a bit more. But I really appreciate having you on. So thanks again. 

52:23
Hammad Chaudhry
Thank you both. Appreciate your time. Take care. Bye. Thank you. 

52:35
Erin Khan
For our listeners, we’re excited to bring you to the leading edge of innovation and construction. Over the course of the show, we will be hosting leaders from the industry as they share their experience and insight. Don’t be shy, please send us an email at podcast@sureworx.io to connect with us. Thanks again and be sure to tune in for our next episode. 

53:00
Erin Khan
I’m Erin Khan and this is the SureWorx Podcast.