Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
E3: How construction projects buy tech
78 Plays3 years ago
In this episode Carlos & Jason chat about how technology is actually purchased in construction, and who holds the power. They also dive into the best way to structure construction data, whether that's location, time or something else entirely.
Recommended
Transcript
Introduction and Preview
00:00:00
Speaker
Hey everyone and welcome to the offsite podcast where we chat all things construction, technology and everything in between. My name's Carlos and I spend most of my days talking to construction teams about how they deliver projects. And my name's Jason and I spend all day building software that construction teams use to deliver projects.
00:00:20
Speaker
So today we're going to talk about image capture technology, the sale of a LECO software, who owns PowerProject, also known as Asta, and Hinkley's latest geospatial portal.
Advancements in Image Capture Technology
00:00:32
Speaker
So first up, image capture technology. It's hard not to notice this stuff. If you get to any exhibition style events, you'll see robots like Spot the Dog, you'll see drone image capture like SenseSats. There's a lot of it at the moment, but in the news this week,
00:00:48
Speaker
We've seen a new, what they call a bespoke camera and buddy system, which basically captures 360 degree views of sites as it moves around, I guess, corridors and floors. And this creates 3D models. Jason, it all looks really cool as usual. What do you think it's actually used for?
00:01:08
Speaker
I find the image capture space super interesting. If you think about the problem, there's different layers to it. So you've got the first bit, which is just being able to capture the site. And that's, I think, not that interesting. Google Maps have been doing it for ages, right? They drive around the stream. When you say capture, you mean literally an image of what it looks like? Well, no. A lot of these platforms have some form of camera on a helmet.
00:01:38
Speaker
or camera attached to something walking around the site and the ability to sort of take captures and then stitch that together into some sort of navigable space. It's essentially like Street View, right? Google Street View. That bit is like, I think, pretty standard technology.
00:01:57
Speaker
Yeah, then there's like the, some people are doing that plus robot slash drone. So automating the process of the capture. And like, yeah, that's interesting. I think like I've seen ones which do like a little crawler robot that drives around. Obviously, as soon as you've got a stair or any uneven ground, you've got a problem.
00:02:18
Speaker
Spot the dog is a regular for like demonstrations of this but I think it's like not there yet in terms of actually being able to like walk a site because obviously in the construction site scenario the site changes every day but the paths are not very well defined.
00:02:33
Speaker
Um, yeah, maybe a building is a bit more consistent, you know, but like if you're talking about, if you're talking about any type of like civil infrastructure, there's, there's obvious limitations of the ability to capture that consistently. Anyway, that's a whole challenge that I don't, from what I've seen is still not there yet.
The Role of Quantity Surveyors and Image Capture Challenges
00:02:54
Speaker
But then there's like this third piece.
00:02:57
Speaker
which is, OK, you've got a situation on your project where you've captured every day this equivalent of Google Street view every day. But what the hell do you do with that? It's nice to be able to go back and look at the data. But how does it actually drive value? I guess it drives value from a claims or having a record if you had a claim to be like, when did something happen?
00:03:22
Speaker
But to find out when it happens, unless you've got, unless you can turn the images into something structured, you have to just go back through time to find when did that change or whatever. So the third piece of this space is like some sort of artificial intelligence, understanding what's been built, and then outputting some data that's structured, whether that's like tracking progress of a schedule,
00:03:49
Speaker
doing measurements of to check that dimensions against the drawings and if you look at the state of platforms in this space it's still pretty early like I would say one of the leaders is like open space and they have a product that's trying to do this and most of I think they can do walls and ceilings maybe or like that's still in a beta and then they've got even earlier betas for things like checking it doorways or in
00:04:18
Speaker
Um, so like it's still a ways off being able to generate that in any meaningful, uh, way. But I, without that piece, it just seems like a lot of captures. It's like, it's a nice to have. Yeah. A lot of it feels like it's like this flashy thing, which is funded by an innovation project. So there's lots of cash to go and do stuff, but like, what is the actual meat and veg at the end of it?
00:04:48
Speaker
You mentioned about the ability to actually understand progress against, I don't know, a bin model or something like that to say, right, this door's been built, this wall's been built. Obviously, QS backgrounds, QS hats on. QS is a famous for never going to site, right? You must know the running joke. They just sit in the office and they'll do anything to not go to site, put boots on and get dusty.
00:05:17
Speaker
Have a drink, I'm sure. Yeah, nice and warm, sit with your coffee. But the big thing for QS each month is they've got to pay subbies, right? So every month you have to pay subbies and you have to pay them based on what they have done and maybe like a forecast of a week ahead to the end of the period. And lazy QS's would go sit down with the engineer and go, right, this list of 10 things, have they done them? And they get an idea about what they've actually done. Good QS's would go out on site,
00:05:46
Speaker
And not to a measure, you're not actually measuring, they call it that, but you're just looking at stuff and thinking, how progressed is this? Shall I pay them? Because if you pay them, then they don't turn up the next day, and they haven't done it, yep, that shit's creep, really. So I could see that being the benefit, but I guess it's going to be this long path of trying to get there, and we're going to get all these gimmicky type things to try and experiment.
00:06:09
Speaker
I think I totally believe in the space. I think that projects in the future will all do some sort of capture. I think the capture technology will depend on the type of project. So obviously like a highway project benefits from drones that are flying every day by building from the workshop.
00:06:29
Speaker
But I don't think any, I really think that the value of the platform is only there once you can turn the images into something structured like progress or quality checking of something. Yeah. Otherwise, you've just got mountains of images to go through. And like you've probably seen like, even on the best projects, like on bad projects, you have very little records of what's happening outside. But even on good projects, you have like lots of images and types of libraries or whatever.
00:06:58
Speaker
But just to go through all that and find it is, it's like too much, it's almost like too much information. So yeah, totally blue in the space, but I just, I think it needs, it needs to get to like this end point or like some progress point because something that gets like real, like I think is really valuable.
Fragmentation in Construction Software
00:07:20
Speaker
I once had to dig out a photo from site and I had to rummage these site diaries, like folders.
00:07:27
Speaker
and obviously every photo is labeled that like IMG underscore 67268 so you literally have to open every single one just to work out where it is or who it is or when.
00:07:39
Speaker
Absolutely don't bet. We can cut this bit or leave it in but we had a chat on Night Shift on a project like 10 years ago and he was doing the site diaries recording the progress overnight and for many months was getting like slightly like people on Night Shift can go a little bit crazy. The site diaries deteriorated to like just like
00:08:06
Speaker
Like at one point he was just taking photos of his boots on the ground being like, these are my boots and attaching that to the site diary or writing messages in there. Like the useless steel fixes didn't turn up again, but no one's going to read this ever anyway. And then eventually like took it another step further and got fired. But for a while it was quite funny. That's why we get so many requests for a site diary app, right? Yeah.
00:08:36
Speaker
Right, so next up we have, and I've got quite a funny note here, it says, Aleko plc slash Aleko software gmbh slash power project slash Asta. Pretty hard to name these guys, but I guess the summary is there's a company that owns a subsidiary, which is the company that provides power projects, which is a common master schedule tool for predominantly the building industry.
00:09:02
Speaker
So this company has been sold to a German company for some of 600k. We'll move on to that valuation at the end maybe. What do you think this means for the users of Asta Power Project?
00:09:16
Speaker
I think it means nothing for us, the power project. So the, the actual story is that when I was reading it is the parent company that owns our project has, it's actually this piece of software that's been sold as like one small piece of like quite dated, like CAD 2D, 3D, um, drawing software, like a really bad version of SketchUp.
00:09:41
Speaker
But like, I think I've got it right. And as I only really recognize this. No, it's not. So if you go into like the, if you look at like, and this is indicative of like the construction, like construction software industry. This is like one piece of software out of 17 that I like to solve. Yeah. It's another, it's like an example of how so many construction software companies are just like vast collections of isolated tools.
00:10:10
Speaker
rather than a single platform. And I think it also tells what is wrong with the construction software space. And, yeah, I think a lot of the time they'll, like,
00:10:25
Speaker
a company will buy a smaller company or a tool or something that's got some traction revenue and then they'll just sort of put it with their collection of other tools and not spend any time or effort to integrate it with the rest of the things that they have.
00:10:42
Speaker
Like if you think about that cost compared to like what the tools we use as a technology business, if any of the platforms we use had some product or part of that offering that wasn't connected to the rest of it, it would be like laughable. But that's the state of play and the norm for like a lot of construction platform. Like as an example, you take Bentley, right?
00:11:08
Speaker
If you go to the Bentley website and then go to the products section and then go view all products, you'll find that they have a product catalog of software that has 12 products per page and they have 12 pages.
00:11:27
Speaker
is like hundreds of pieces of software that Bentley owned. And they're just acquired and acquired over time. And so little of it is actually connected with other pieces of it. And that's why things like Procore that are taking a proper modern approach to a single platform where things are connected with each other. And if you release a new module, the RFI module talks to the site driver.
00:11:55
Speaker
that's just a completely different approach to the way that it is in the construction industry and like as an industry how do you expect construction companies and contractors to make sense of 144 products right number one and then like number two how do you expect them to adopt such like a disjointed mess of rubbish so yeah yeah like are you
00:12:22
Speaker
Do you have sales team for each? Are you selling a bundle? Do you then onboard each one? That's a nightmare, right? Well, not to mention that many of these people outsource their like selling function to like partners.
00:12:37
Speaker
So you end up with like a third party reseller going, here's a catalog of 144 pieces of software alongside the other, all the other vendors worth of software in their catalog, which ones do you want? It's like a, it's the worst buying experience ever. So I don't really have a thought on the sale of this software. It seems like a fairly small dated pieces.
00:13:07
Speaker
software. Um, but it, it, it did roll me up about construction software in general.
Geospatial Data Integration at Hinkley Point C
00:13:14
Speaker
And like, what a mess it is. Yeah. I never, it just, it reminds me of the old stuff we used to use on projects 10 to 15 years ago. And it's sort of just sat there with that developed. The thing that was stood out for me though, was they sell that part of the business to 600 K a business that claims to have a hundred thousand users.
00:13:35
Speaker
That seems wild to me, like there must be something. The software that they sold had 100,000 or the parent company did? The subsidiary, the Aleko Software GMBH, the subsidiary that they sold to the German company, that set of software claims to have 100,000 users and they sold it for 600,000 euros. So something's not right there.
00:14:00
Speaker
I don't know if you looked at it, but it's in a category where there's like massive, it's in the 3D model. There are so many good offerings coming onto the market. You're going to spend millions in development. Well, you'd basically start again. But yeah, I don't know why someone buys it. I've never heard of it before. Switch off it. Yeah, yeah.
00:14:31
Speaker
Right, so Hinkley Point C, hard to read the news at the moment and not keep coming across Hinkley Point C, there's obviously a lot going on. This week there's an article detailing EDS enterprise GI system which is developed and delivered by Esri and it's a geospatial portal for the project and they claim to sort of pull together things like spatial maps, maps, apps, dashboards, all of this information into one sort of
00:15:00
Speaker
hub to give everyone access to this information. Jason sounds simple and great but I guess what's the reality behind this and collecting that sort of volume of data to put it in one spot. I guess I'd be interested to know your thoughts on this. You spend a lot of time every day talking to teams
00:15:25
Speaker
integrating like GIS tools with our software and you've got more exposure to how teams are using it so I think yeah I'd be interested in what your take is. Yeah so whenever we speak to the sort of larger projects and that is generally the larger ones that have systems like GIS it's not really found on small five million pound jobs
00:15:49
Speaker
you end up with lots of different departments who never speak to each other, but all own their own sort of data set and are feeding one part of the project, like in their own little world. So firstly, we say we have a meeting with the GIS team. It's like they've never met anyone outside of their dusty little office before. It's quite like a siloed set of teams. So
00:16:15
Speaker
I can imagine that sort of job, you're going to have tons of consultants who are basically guaranteeing work for the next 10 years because it's not this sort of, from my understanding and my experience, it's not this automated flow of lots of data into a bucket and everyone accesses it and it's nice and seamless and easy and done. There'll be massive teams behind it actually pulling this together. So yeah, it's
00:16:45
Speaker
The output sounds great to actually have a hub that pulls all this stuff together. So then we're turning what could be eight parts of information, for example, into one. But how up to date relevant and seamless the data is and the sort of process of actually putting together and keeping it current. I have no idea.
00:17:07
Speaker
like, like the illusion of something automated, but behind the scenes, there's like, yeah, people on treadmills, like making it work. Yeah. We've, we've, we've spoke to a couple of projects recently and just to get like drone imagery into affects because obviously we pull through that data. There's like six guys that have been committing X hours per week to convert information or the, um, the data types. And it just seems nuts that it's not just
00:17:34
Speaker
Yeah, an API to set up and it's done and it's there and that's feeding forever. The what's, what's super interesting is like going back to what we were talking about before with the, the like reality capture thing.
00:17:48
Speaker
And, you know, we're talking about hunting back through images of the site, like, site diaries and trying to work out where the information is. In construction, the concept of, like, structuring information geographically as, like, that is the primary way to get to data makes a bunch of sense and is super, super valuable.
00:18:12
Speaker
It breaks down so many barriers. You can lay out so much other information over that. Like you could have scheduled performance, you could have plans, you could have the design, you could have the eyes built records or the reality capture.
00:18:32
Speaker
like if it's geographically structured, the retrieval of that information, like one of the massive problems with getting to the information is like retrieving the stuff that you actually need. Like you want timing, location stamp basically, don't you? So it's all the right time.
00:18:49
Speaker
having stuff geographically organized and accessible in a platform is massive for information retrieval and construction. But, and then this kind of ties together like the three things we talked about actually, because yeah, so having the information capture
00:19:08
Speaker
that links to the importance of geographic organization of data for retrieval, but then the reason it's hard to get to that is because so many of the tools that people use for different bits of the information are like the standalone products, right? So someone might have a
00:19:27
Speaker
cost-value reporting system that's done in a desktop piece of software that needs to import a spreadsheet from something else and then outport a spreadsheet over here. So to get the information actually in this one, like, so they're using Esri's system. That's a fairly, like, accessible platform. We can integrate with it, like, quite easily. But if you've got a piece of desktop software or something that runs in a spreadsheet sat on someone's computer on their C drive,
00:19:55
Speaker
in an office somewhere, that's not accessible. So then this proliferation of, or this historical set of tools that are not connected with each other, make the putting that information together in one spot really hard for contractors. Yeah, definitely. And then also just, you've got different teams and reports driven off different
00:20:24
Speaker
datasets when it's not pulled. So like on Crossrail, within project controls, there's this report called QURR, which is like a quantity unit rate report. And it's basically how much have you done times a rate times, so the rate, how much you've done. And then that works out progress against things like concrete, against the WBS, which then drives the overall value from the project.
00:20:51
Speaker
And because there was no method of capture for that, the project controls person would go do a lack of sight and write down stuff. So they're doing that. You've got the QS is going out, looking at stuff, deciding how much they're going to pay. And every, every single department has got a completely different number. Nothing ties up. And then
00:21:07
Speaker
You spent three weeks arguing about whether the data is accurate rather than what it actually means. Yeah. But if you think about then, if you apply the fact that there's so many disconnected, isolated systems, maybe the QS goes out and estimates how much was done. But probably for some other tracking spreadsheet somewhere, the engineer is probably doing the same thing. And then probably somewhere else, a quality person is doing the same thing.
00:21:35
Speaker
So like it really just, if there were, if there was a single platform or, or it just goes to show the power of like having something like a platform that's in the cloud, because the data can move between systems, uh, automatically.
Integration Challenges in Construction Data Systems
00:21:52
Speaker
Yeah. Cause it, yeah, not only is it like such a ridiculously manual process, but it guaranteed it's being done three times, you know?
00:22:05
Speaker
Yeah. And every, each of these vendors are going to be pitching these productivity games, but they're, they're individual. What game are you really going to get? It's, uh, yeah. Yep. Right. Just to like close the loop on that, right? Like you think about even I just think about like, um, our product and like scheduling. Yeah. We deal with like helping engineers build their schedule.
00:22:34
Speaker
But if you think about the other spreadsheets that engineers complete, like the amount of things they have to do, they'll have to complete some register of like out of hours working for like a community team. They'll have to complete a schedule of like a concrete demand for like a concrete batch plan. And they'll have to have a spreadsheet after spreadsheet, ones for like a safety team of like high risk activities.
00:23:01
Speaker
all that information is duplicating information that would be in a schedule and so like one of the things that we've intentionally done is made it so you can just attach that information to your schedule and then it can be output in in a format you do it once and yeah there's just a million examples in construction of like the same thing being done three times. Yeah definitely and it
00:23:27
Speaker
to detach the examples you've just given. So how much concrete you need from a batching plant or how much maybe you need to detach any of that from a schedule, as we say, crazy, right? I don't know how things can function without the sort of time element attached to it. That's everything. In construction, everything should be attached to both schedule and the location. And I think if you do that, you eliminate. So you eliminate like so much busy work.
00:23:58
Speaker
having everything attached to a geographic location and a schedule. If your cost forecasting was properly attached to the schedule, like properly attached not, there's a schedule with some reference back to it so that whenever the planners eventually once a month update the schedule will get properly attached in a system where they talk to each other. Yeah. And the kind of where being wanted to go potentially is going.
00:24:28
Speaker
But the problem I was found with BIM was it was always three or four months behind. It's like this retrospective thing that they're sort of pulling together behind rather than it actually being used to look forward or plan.
Critique of BIM's Current State
00:24:41
Speaker
That could have changed in the last five years since they've been involved heavily. But yeah, that's what they should be doing. If you're going to slag off BIM, stick with it. Don't back away from us. Yeah, we all know what happens with BIM anyway.
00:24:58
Speaker
Right, that's all we've got time for today. As always, thank you very much for listening.