AR and Autodesk Construction Cloud: Powering $9B Project Excellence

WALEED ZAFAR: Hi, everybody. My name is Waleed Zafar. I'm the director of Mission Critical here at XYZ Reality. And thank you for joining this session, where we're going to cover AR and Autodesk Construction Cloud-- Powering $9 Billion Project Excellence.

So to kick things off, I have a short video explaining who we are and what we do. But a bit of background about myself-- I actually joined XYZ Reality six years ago. And I actually joined XYZ as its first employee.

I actually met David Mitchell, founder and CEO, six years ago. I was actually a student at London Business School at the time, completed my master's when I bumped into this tall, crazy Irishman who told me about this vision he had of builders building from holograms and getting rid of 2D drawings. And I thought, you know what? This makes a lot of sense. So I ended up joining him as the first employee.

Since that point in time, our business has grown and expanded. We've raised close to $50 million in funding to date. We've deployed our technologies all across the world. And we're now a team of around about 130 people as I record this information. So to kick things off, here's the video.

[VIDEO PLAYBACK]

[SUSPENSEFUL MUSIC]

[END PLAYBACK]

So what are some of the challenges that our construction industry faces today? Well, we summarize this into three themes speaking with our client and partners as of today. And first one is around delivery overruns. And I think this is one that probably everybody can actually relate to, which is the fact that more often than not, these project sites that are kind of large and complex in scale-- more often than not, they're actually delayed when it comes to, actually, the delivery of these assets.

Now, when you try to understand the reasoning why, what's driving this issue, oftentimes it's actually down to the labor shortages and the market today. And it's not just labor shortages in terms of the boots on the ground. But it's also the management here, too.

And if you think about what that means, essentially, you're having to promote inexperienced individuals into experienced roles where they actually have to deliver these complex projects without actually having the necessary tools that enables them to do so. And all of a sudden, what that results in is that you might have one project and one particular area that's gone fantastically well. But on a different job site, it's been a complete disaster. And sometimes, this can even happen within the same GC itself, and the same supply chain being hired as well.

So how can we help prevent these delivery overruns? How can we help bridge the gap in the labor market? And how can we actually provide that consistency in quality delivery?

And speaking closely with the Autodesk team, as well, and in terms of their solution and their offering, really, what they're hoping to be able to achieve when we had our discussion about our partnership together was being able to increase the engagement of their solution. How do they also improve the satisfaction of their solution on project sites and, therefore, then increase the retention of these users for the next project sites and actually increase their user base as a direct result of this, all supporting the delivery of these project sites?

So we conducted a survey with all of our clients and partners in our network. And 99% of our construction leaders within our network actually stated that they delivered projects out of schedule within the last several years, which I think probably everybody here can relate to.

Now, what we actually also found out is that 30% of all construction activities are rework activities. And so if you think about what that means, essentially, what that means is that 30% of the human capital time on any given project is allocated towards fixing issues that could have been built right the first time. And herein lies the biggest opportunity that we see for improvement in our industry today.

And so to conceptualize this problem, if you think about what happens today in this industry-- is that we spend all this time coordinating and designing this perfect 3D model, making sure it's completely clash-free. But when it goes out to the field, we actually have to convert this 3D model into thousands of 2D drawings in order to be able to construct this 3D asset.

And if you think about it, 2D is not a natural language for us humans. We're 3D creatures. We operate in a 3D environment. Yet we task our trades today to look at a 2D drawing, conceptualize a 3D asset from that 2D drawing. But the most challenging part is actually being able to position that information out on-site with millimeter accuracy.

So a bit more background about us is that when we actually started off business, we actually set up our own laser scanning subsidiary. And we were actually laser scanning and providing site survey services to some of the largest projects in Europe, including hyperscale data center projects. And what we actually found was that more often than not, these were fundamentally reactive processes.

So even if you take other reality capture solutions, like 360 cameras, and we already mentioned about laser scanning and LiDAR scanning, et cetera, from there-- if you think about what happens-- is that you have to wait for something to be installed. You then scan that information, scan that data.

You get that point cloud model, let's say. You have to clean that point cloud model. You overlay it against the design. And only then do you see the deviations between what's been built versus what's been designed for, at which point in time you have to make one of two decisions. One, do I update the model to reflect what's been built or, two, do I go through a rework procedure in the field?

Now, more often than not, you're probably going to do the former rather than the latter. But you can only update the model so far and so much until something critical has to be installed. And it then no longer fits. And that's where these unforeseen rework expenditures really and truly come from, the unforeseen delays that happen on project sites come from.

And so if you think about this, we're constantly reacting to situations. We're waiting for something to be installed. And only then do we check it. And you don't find out about these problems and issues up till a week to two weeks, at times, at which point then it's not like construction stopped waiting for this verification report to come through. Construction progresses. So being able to resolve that challenge there is of the utmost importance to us.

And so just to provide you guys with a bit of data from our time as a laser survey business, well, when actually doing our spot checks, we were actually only able to spot check about 2% to 5% of all installations in the field. And we scanned everything out on-site. But yet, despite this, 30% of all activities were rework activities in the field. This is just reinforcing that reactive nature. So how do we help shift the industry away from a reactive approach to a proactive approach?

So to summarize the challenges that this industry faces today, being able to de-risk that project delivery, how do we remove this lag of information as to what's happened out on-site versus what's being reported on? How do we verify the accuracy of these installations in real time? How do we eliminate any of the subjective data that comes from the field?

As of today, in the best of cases, is someone walking out on-site with an iPad with the drawings, looking around the room, and going, yeah, we're about 20% complete here? How do we make this a completely data-driven and completely objective approach and, therefore, then eliminate any of the disputes and rework that actually occur on project sites today?

So this is what we see the solution as-- being able to design in 3D, build in 3D, and validate in 3D. And this happens through the Atom and the XYZ platform solution.

The Atom itself, the headset, is the world's most accurate augmented reality headset. And it's also the world's first engineering-grade augmented reality headset. And essentially, what we're able to do is position hyperscale BIM models out on-site with millimeter accuracy.

And it also has its first of its kind integration with Autodesk BIM 360. And I'll actually walk you through a case study of when we actually launched this live midway through a project. And some of the results were incredibly impressive.

[VIDEO PLAYBACK]

[ROCK MUSIC]

[END PLAYBACK]

So a bit more about the Atom-- well, it's a fully safety-certified hard hat and visor integrated in to the Atom. So it complies with all the health and safety regulations in construction today-- so EN397 for the hard hat, EN166 for the visor. And in the North American standards, ANSIZ87.1 and Z89.1.

And essentially, from the headset perspective, it actually has 32 gigs of RAM and a terabyte of storage. So it's really, truly enabled to be able to load up those large-scale, hyperscale models out on-site with that millimeter accuracy.

And from a security standards perspective, we comply with SOC2 and ISO 27001. Our information as of today is able to be stored on Azure Cloud. But we've also built our entire cloud platform to be agnostic such that if you require for it to be spun up on, let's say, GCP or, let's say, AWS, that's no problem for us as well.

Furthermore, the battery life on the sensor is around about two hours. And these batteries are hot-swappable-- so similar to a drill gun, if you will. So the hot-swappable batteries, which mean our users can go out on-site and spend a whole day in the field-- six, seven hours a day out on-site with the Atom--

So we touched on there. And we showcased through the video, too, about the integration that we have with the Autodesk ecosystem. So we're actually able to export the models directly from Revit and Navisworks. We actually have a little plugin into both that enables you to be able to, through one click, export the models to our cloud platform. That then downloads into our headset, enables the user to be able to go out on-site and view the model in context.

And furthermore, too, when we're raising issues in the field, that then feeds back into BIM 360 and the ACC suite such that not only are we just raising an issue, but we're actually raising it in the direct XYZ location of where that problem occurred such that you no longer have to try and understand where this problem was on the job site. Everybody in the office has that contextual information to had, including the image attach of the problem itself.

And actually, when we deliver this service, we actually deliver it as a fully managed service. Now, why do we do this? Well, typically speaking, on project sites today, you and your teams are already at 130% capacity. So you throw in a new piece of tech and ask the team to figure it out, integrate it in. The chances of it being successful is probably slim to none.

So what we actually do is provide the boots on the ground, the engineers who are actually going to go out on-site with the headset and ensure all the works that are being completed in the field are being completed as per design. And they're actually tracking the progress of all of this for you guys. So you guys get the value that you need without having to do the heavy lifting. We take care of that part for you.

And essentially, what we're able to do is actually change cultures on project sites today. We're actually able to shift project culture to take a more proactive measure. If you can see what you need to build, you're going to build it right the first time. Build it right the first time-- you eliminate rework and clashes from your project sites today.

And so across all of our deployments today, 64% of all the construction issues raised on project sites have actually been raised through the Atom. So if you think about what that means, essentially, almost 2/3 of all the information being received by project teams today would never have been caught without the Atom being present. And that right there is an extremely powerful piece.

Being able to provide that level of transparency in real time enables construction teams to be able to action accordingly in real time. That means that issues get resolved faster. They have all the information they need to be able to resolve the issue. And therefore, then the project is able to progress without experiencing those unforeseen delays and unforeseen rework activities on job sites.

And so just a bit of data in terms of how the split of the Atom is on-site-- essentially, when our engineers do the inspections, this falls into three different buckets-- the pre-installation inspection, the during the installation inspection, and then the post-installation inspection. So if we take the Atom out in the field, 13.1% of the time, it's likely that it's actually doing pre-installation inspection-- so visualizing the next set of work to take place in an area to ensure the constructability of the design.

So for example, if there's any hard clashes that are about to take place, we're actually able to identify them before the trade actually comes into the area such that the design team can make the necessary changes such that the trade, when they actually arrive on-site or arrive into that particular environment or that particular area-- that they don't have to spend times with tools down trying to figure out how they're going to actually resolve the challenge. That part's been done ahead of time.

58.8% of the time, we're there during the installation. What that means is that as the works are taking place, our engineer's there, ensuring whatever is in the model is, in fact, being constructed out on-site. And so that provides that real-time validation. And therefore, then if you check works before the installation and during the installation itself, well, then you don't have to check that much works post-installation.

So only 28.9% of the time are we actually checking works after they've been installed because we've verified all the installation in real time. And all of that results in less than 1% of all construction activities being rework activities on the project sites we've been deployed on. And this 29% reduction in rework activities has been extremely powerful and extremely valuable for our client partners today.

Now, one of the things that we've launched in the last year or so has been our project control solution. And walk you through how that works-- is that there's three steps to it. There's this capture piece. There's a process piece. And then there's the data piece.

So to kick things off on the capture piece, how that actually works is the following-- that our engineer goes out on-site. And before they go out on-site and before we actually come onto site, we actually take the models. And we actually take the construction schedule. And we actually link the two together, therefore, then essentially creating a 4D model. So when the engineer goes on-site, they're actually validating everything that's been installed, ensuring the quality of its installation, but also confirming it's, in fact, been installed in the field.

So to show you what that looks like, I've got a little video here in one second. So this is actually live from one of the project sites. And so what you're about to see is a POV of a user.

So they've loaded up the menu. They've gone into the models, gone into the structural section. And they've just selected now to see the structural models in context. And we can see here the yellow overlay against the steel columns that have been installed. The go-ahead, go into our progress feature-- and for the purposes of this demonstration, we deselected a few elements that we have already been installed just to demonstrate how it work. And what we're going to do is simply just highlight a couple of the assets in the model and just select them as being installed. And it's literally as simple as that and being able to mark works as being installed in the field itself.

So we've essentially created the world's most advanced data capture solution in the market today. So we're capturing everything that's been installed. We've verified the fact that it's been installed correctly and within construction tolerances. Now, where does that data and information go?

Well, essentially, then we've also created a model viewer platform such that we can actually process all this data and information. So everything you're seeing in green over here has been installed. And it's been verified as such. That's why you're seeing it in green.

Everything in red over here has been identified as late, meaning it's not installed yet, but it should have been installed as per the construction schedule. And we can see this through the little widgets that are here on the side about steel erection, civil works, cladding, roofing. All of this data and information is totally customizable to your project needs and requirements.

And the most valuable piece actually comes in the data side of things, too, such that you're actually able to track your projects across your entire portfolio such that you can get a quick, high-level view on the cloud platform itself. Click into any one of the projects. You get a nice, high-level-looking dashboard. You have your exact percentage of completion on the top left, that 52.68% completion.

And then when you're looking at that S-curve, that yellow line is actually your baseline program-- cumulative baseline program, I should state. The blue line is actually the latest program that's come from the GC. And the green line is the actual installation that's taking place out on-site.

We've also brought in some of the 4D model snippets. So for the hacks, mechanicals, and electrical work-- so everything in red you're seeing there is late. Everything in green has been installed. And everything in yellow-- these are the planned activities for the coming weeks.

We also break down this information by the main packages of work. So you can see it for the CSA, the mechanical and the electrical works, and then also then for the sub-packages of works, too. And we also bring in on the right-hand side over here the exact percentage completions based on the main packages of works against the baseline program, the latest program from the GC, versus the actual installation. So it's real-time, objective data that is being published every single day for your project teams to be able to ask the right questions and know exactly where you're at on the project site itself.

And then for the project team, they can delve deeper. So they get the overall looking S-curve-- so the yellow, again, being the cumulative planned installation and the green line being the actual installation. So in this example here, there's a bit of a delta between the two.

Click into that. You get it broken down by the main packages of works-- CSA, electrical, mechanical. And let's say, for example, you want to break down the CSA further. Then you get it by the sub-packages of works-- structural, concrete, cladding.

Let's say, for example, we want to understand why there's a bit of a delta between the structural works. Click into that, and you get the 4D model snippet. So everything in green that you're seeing here has been installed. Everything in red should have been installed by now, and it's not. So it's late. And everything in gray in this particular case is actually the planned activities for the coming weeks. And all this information is in real time.

This is the ground truth that are actually-- is actually happening on this particular project. So in other words, it's a mirror world reflection of the actual job site itself. And you get access to all of this data completely remotely.

So I mentioned there about the BIM 360 case studies. So when we actually launched the integration with BIM 360, we actually did this in partnership with PM Group and Autodesk.

So why? Essentially, from the customers' perspective, it was trying to improve the Autodesk ecosystem, increase the return on investment for investing into these digital tools. And how do we add more value to the digital environment overall such that they get a more effective solution in managing the quality and the project controls on their project sites today? They can then actually de-risk the project delivery, increase the quality on the job site, reduce the rework, and actually get some real-time data. So there's no longer subjective information.

From an Autodesk perspective, they wanted to be able to better improve their daily active users and monthly active users. How do they make sure that they actually get engagement from those outside the BIM department, if you will? How do they get better engagement to the package managers, to the superintendents, to the project directors, et cetera, as well?

And then from an XYZ perspective, we wanted to get a much wider network effect taking place. We wanted to develop a product with a leading brand and actually then demonstrate how we could actually enable further integrations and become a trusted partner of Autodesk itself.

So essentially, then what we did is that we created that integration such that whenever our users were out on-site and they saw an issue, rather than raising the issue, taking the image of the issue, coming back, and creating some of PDF, how do we get that to actually link seamlessly with the models on BIM 360 and the ACC platform such that at the moment that someone raises an issue on-site, it automatically attaches to the federated model in the exact XYZ coordinate point? And this is, in fact, what we're actually able to do for this project site. And so with this particular project, huge hyperscale data center project in Europe-- and client was heavily involved with actually tracking this project, too.

Before we'd launched the BIM 360 integration, 38% of all the issues that are actually on the job site were actually created by XYZ-- the 38% level 2 construction issues, that is. When we launched then the BIM 360 integration, what we actually saw was an increase in the number of issues out on-site. In other words, we actually created a new flywheel effect. We're actually able to report issues faster, which then meant that PM Group were able to action them faster, which then meant they got resolved faster. And then the project site could move along to the next phase of works much faster.

So this flywheel effect that took place was apparent seeing, actually, the data itself. So before the BIM 360 integration, we raised, on average, about 47.5 issues per month. Straight after, 145.1 issues per month-- this 3x increase in issues raised in the fields was incredible to see on the job site itself.

And again, this is just showing you the month-on-month information, the month-on-month data. And it's really clear to see when the BIM 360 went live and the powerful effects as a direct result of this.

So on this particular project, we came through about midway through the job. The project was in distress, it has to be said. And the reason why we actually came onto this project site was that an EOT claim has actually been placed in by the GC to the client. And it was stated that the copper went missing on site. And the client requested us, through our system and our data, to actually prove out the fact that they weren't even ready to receive the copper, let alone actually using that, as a reason why they need an extension of time.

So what we did-- we caught up with the data. The GC was claiming they were four weeks behind schedule. And we actually, in fact, proved that there weren't four weeks behind schedule. They were, in fact, 14 weeks behind schedule. And we proved this through the models, through the 4D data, and all through the XYZ platform. And we actually protected this client's interest. We actually saved the client $6 million worth of EOT claims, and also prevented further damages to take place with their customer, also-- so actually saved them $3 million in liquidated damages.

Now, in this particular project, what ended up happening over here was that there was a permitting issue on the job site. So it actually delayed the kickoff. Now, I'm sure this is something that everybody can relate to, that permitting issues happen on every job site. And despite that, the customer still wants to get their asset on time.

So once the job finally kicked off, we actually worked very closely with the client and the GC in being able to support an accelerated delivery program. And this is what you're seeing over here, this huge spikes in productivity that happened towards the back end of the job. And this all happened through to-- as a result of the XYZ platform.

The project team having that real-time data and information, assessing exactly where they're at at any given hour, every single day, all throughout that project duration-- and this is what actually assured the client that they were actually going to be able to deliver the asset on time to the customer. They were able to assess the performance of their supply chain in real time and, therefore, then incentivize the supply chain and see the results of the incentivization to their customer. And as a result, that customer got the asset on time.

The last thing on this particular case study-- so this one's actually a quality case study over here. What ended up happening was that there was actually going to be a clash in the busbar installation. And so we could foresee this through the Atom headset.

Now, this particular issue was actually on the critical path as well. Now, it wasn't due for installation for another six weeks. But what would have happened if we weren't able to spot this problem, this future installation clash that was about to take place, was that they would have-- the trade order package would have come on-site.

They would have seen this problem. They would have realized that they can no longer fit their current installation plans. And they would have actually had to have ordered new parts. And those new parts to arrive would have taken six weeks. But because we were able to foresee this risk taking place, we were able to flag it up in real time and actually save the client a six-week delay on that critical path and actually save the project $337,000 as a direct result.

[VIDEO PLAYBACK]

[ROCK MUSIC]

MALE SPEAKER: In construction, the stakes are high. There's no room for error. Accuracy shouldn't even be up for discussion. Some seem to think close enough is good enough. But we don't have the tolerance for second-best.

Engineering-grade augmented reality means millimeter precision, the only solution to protect your bottom line because accuracy is everything.

[END PLAYBACK]

WALEED ZAFAR: I hope everyone found that session informative. If you do have any questions, please feel free to reach out anytime. Thank you.