Feb 13th, 2020
Engineering Influence welcomes Adam Klatzkin, who is vice president, business development - iTwin Services at Bentley Systems, to talk about digital twin technology and its accelerating impact on the engineering industry.
Host: Hi, this is Gerry Donohue with ACEC with another episode of the Engineering Influence podcast. One of the most exciting technological developments in the A/E/C space in recent years has been the digital twin and Bentley Systems is a leader in this field. To learn more about it, we're talking today with Adam Klatzkin who is Bentley’s Vice President, Business Development - iTwin Services. His team is responsible for the firm's digital twin cloud services for enterprise infrastructure engineering. Welcome, Mr. Klatzkin. How are you?
Adam Klatzkin: Great, thanks for having me, Gerry. I'm excited to be here today and to talk to you about digital twins.
Host:Great. So, so to give me and the listeners a firm foundation into what we're talking about, what exactly is a digital twin?
Adam Klatzkin: So to try to very simply describe it as a digital asset that you can trust represents a physical asset as it exists today. And typically they are joined together and kept in sync. So you can leverage the digital twin to understand what's happening with the real-world assets in real-time; to have visibility into the real-world asset without having to leave an office; to collaborate easier around it and more effectively with distributed teams or stakeholders; and ultimately to make more informed decisions that lead to better performance. It might be a better project performance if you're designing or constructing the asset. It might be better operational performance if the asset already exists. But that, that's the end goal and the value to come out of it. I'll also mention that of course as a twin that relationship is in and should be symbiotic between the digital and physical assets.
Adam Klatzkin: You can imagine this continuous feedback loop of sorts. The analysis and insights that happen within the digital twin become actions and interventions that feedback to the physical asset and result in outcomes or changes the feedback to the digital twin. And that cycle just continues. And there are lessons to be learned in mind and the impacts of decisions that are made along the way can be digitally captured. So ultimately an organization that leverages digital twins is not only going to reduce costs and risks for a single asset, but that digital transformation can allow them to optimize performance across an entire portfolio and future projects as well based on the lessons that have been learned.
Host: Well, that's, that's quite comprehensive. To take it down to just a single project, for example, a bridge, how would an engineering firm use a digital twin in the design and construction of a bridge?
Adam Klatzkin: So that's it's a great question, but to answer it honestly, I'd like to start with kind of the ugly truth and challenge some of the ugly truths and challenges of an infrastructure digital twin for any type of engineering project--a bridge being one--to better understand what you would need and how you would work with it. So the concept of digital twins isn't new. It has been used in manufacturing for years, but why haven't they really been adopted in infrastructure? One ugly truth is the data. Engineering projects tend to have massive amounts of data. We've seen bridge projects that have multiple design tools in play by trade or discipline. There are modeling tools, geotechnical, structural, beyond just the design tools. There's are tools for and data for analytics and simulations, scheduling, resource management. The list goes on.
Adam Klatzkin: All of that data is siloed. It's all typically understood by a single application or expert users and not available in other contexts. Even worse, engineering projects have tons of change in them and that change is relentless. It's the nature of an engineering project. If there wasn't change happening, then the project isn't progressing. So that's one ugly truth. The data is a mess. The second is, no one vendor could ever provide a digital twin that was locked into their ecosystem. If we consider that bridge example and I said there are many different tools in play, we've seen many projects that use a combination of Bentley products and our competitors. There may be AutoCAD Civil 3D, Bentley OpenRoads, Bentley OpenBridge Designer. All may be used on the same project and others as well.
Adam Klatzkin: So even get started with it a digital twin or even get started a digital print platform is needed to address those ugly truths. And that platform needs to be able to align and federate data together with all this data coming from those many different vendors. And the platform can't just handle the relentless change but it has to leverage it to provide better change management and insights based on the impacts of changes that are occurring. So Bentley has introduced a platform that addresses these challenges, iModel.js. That's our foundational development platform and iTwin services, which is our commercial offering built on the platform. And our philosophy around this is openness, open in every regard, open to any tool by any vendor, any repository where data may reside. And we've even open-sourced the platform itself.
Adam Klatzkin: So to come back to that example of the bridge digital twin platform, addressing those challenges and context to a bridge. You would consider that they'd be very early on in planning. The most important thing to understand for the engineering team through the twin would be the physical location, the site where that bridge is going to be designed, engineered, ultimately constructed. So reality capture could be used via drones to establish a 3D-mesh. That's one example. There are other ways to get reality data too. That could be combined together with map data, terrain data, and all brought together in the context of a digital twin that is the basis and leveraged for planning. This gives anyone at any location access to the true site conditions. Those drones can be flown as often as necessary to keep that digital twin. That's just a reality model right now, continuously surveyed and synchronized. And now as design and engineering progresses, the files that are created from design tools today, they'll likely reside in a connected data environment like ProjectWise, Bentley’s offering there.
Adam Klatzkin: And connections are established between the digital twin and the files in the connected data environment, the design files, through things that we happen to call bridges, an overloaded term in this context, but they are also the bridges that bridge data from the connected data environment into the twin. They understand the native formats that are in play in all those silos. They detect changes as those changes occur. They make those changes available in the digital twin and they align all of that disparate data together into a single digital representation of that physical bridge that you're engineering. And as these changes come into the twin, because as I said relentless change, as they come into the twin, the twin itself can validate the trustworthiness of that data through data quality services. And this is all automated, validated against codes and standards and whatever other rules the engineering team might want to dream up as well.
Adam Klatzkin: Immediately the engineering team can start recognizing the benefits through workflows such as integrated coordination and design reviews where they've got all of the richness of not just having the mirror of what they're designing against the physical reality, but understanding what has changed across all this disparate data that's been aligned together. What changed? When did it change? Why? Where? They can do clash detection against that continuously up-to-date twin and they can do issue resolution. So you get all those benefits in easy web-based accessibility without requiring any installed software. They can make it available to clients for communication to the public stakeholders or whoever else may need access as well.
Adam Klatzkin: And then they can start to introduce connections to other systems and information, like for instance, maybe there's a risk register or database in play, hazards and such, that they can connect that information into the digital twin and have it for visual sharing and, and recognition. They can run simulations against the digital twin. They can run insights and analytics to understand, "Okay, with these changes that occurred over the last period of time since our last coordinated design review, what's the impact of those changes on our costs, on our schedule, on risks," and have those dashboards recognized through the twin. But most importantly, as they go through this entire process, it's not intended to be disruptive. Those engineers and designers are still using the tools and the systems that they use today, they get the added benefit of the twin at being complementary to those systems where everything's being brought together for those integrated workflows.
Host: And just so I understand, right? So if the contractor is using a different system it would automate whenever they input something into their system, it would automatically move over into the Bentley iTwin system and be available to everyone else on the team
Adam Klatzkin: That is absolutely correct. Of course, it does have to be configured that way. But that's correct.
Host: So there's really no extra work. There's no extra inputting. Anything you input into your own system if everything's configured correctly works its way into the twin automatically.
Adam Klatzkin: Not disruptive. No extra manual intervention. No duplication of data.
Host: So you mentioned earlier that one of the reasons why digital twins haven't really made it deeply into the engineering industry so far has been just the huge amounts of data. How far and how deeply into the industry have twins come? Have you a lot of clients out there? A lot of potential?
Adam Klatzkin: There are a lot of clients, a lot of potential. There is no doubt. It is an emerging technology. Anecdotally I can tell you there isn't a single firm that I or our teams here I've spoken with, that isn't exploring or implementing their digital twin strategy. But there's also some interesting third-party research showing that the digital twin market in very short order is actually going to surpass that of BIM. Now, that doesn't mean BIM is going away. It has incredible value and that the market is going to continue to grow at a respectable rate, but digital twins don't replace it. They complement it. They advance it. And they enable further digital transformation. The use of cases, workflows, benefits are much more unbounded in a digital twin than they are in BIM, which is a static file representation at a moment in time.
Adam Klatzkin: Those projections and that research show that in 2024, the BIM market size will be $13.2 billion in U. S. dollars with an 18% growth rate. While in 2025, they're skewed by a year, the digital twin market will be $29.1 billion with a 37% growth rate. If you look at the charts, it's amazing where they intersect and the rate at which digital twins continues to grow. As for today's adoption and initiatives that are out there, again, this independent research, not by Bentley, is showing that of all the digital twin initiatives that they've surveyed, about 16% of those are being driven by engineering and engineering firms; 33% of them are being driven by operations and owners; and the remainder of those initiatives are being driven by supply chain, finance and others.
Adam Klatzkin: So it is definitely skewed right now towards owners and operators driving it, but there's a significant drive within engineering services firms directly. Leading the way by industry penetration of digital twins is transportation. They have the most, again in the same survey, they have the most digital twin projects that are implemented, budgeted or planned within the next three years. Transportation is definitely a leader and that seems to be followed closely by plant design, particularly oil and gas and chemical and then metals and mining.
Host: All, all those things you mentioned are big, big projects, a big infrastructure project, a big factory project. Is there some sort of size limitation for when a digital twin is effective or either affordable or can you do it on a small project as well as a big project?
Adam Klatzkin: From our perspective, the commercial model scales. It's not an inhibitor for any size project. And the benefits certainly apply to all ranges as well. But as far as penetration in the market, those were the areas that seem to see the most activity underway, the industries and types of assets that see the most activity underway right now.
Host: When I was reading about this in the stuff that Bentley's written about digital twins, a statement that came up that sort of struck me was that a digital twin is a connection of data sources. Would you talk about that?
Adam Klatzkin: Or, yeah. If we go back to those ugly truths that I mentioned earlier. A project or asset may have different sources of truth for information. That source of truth may be some bespoke database, it may be an enterprise database and maybe a design file. It may be a connected data environment involved that is a collection of those sources of truth of information. A digital twin should not be a source of truth. It's a view of truth. You can also look at it at a single pane of glass. That's another descriptive term we like to use. The digital twin needs to be a federation of all the relevant information and data sources wherever they reside, brought together into this single pane of glass. It would be a losing proposition to say that data needed to move somewhere else in order to create a digital twin and, and we chatted about that earlier in this discussion as well.
Adam Klatzkin: Generally, we're talking about three types of data sources across an asset life cycle and the connections that might be formed: ET, IT and OT. Data coming from engineering technologies, informational technologies, and operational technologies. The digital twin needs to federate all of this to make information easily accessible in that single pane of glass for immersive visualization, for analytics, for improved decision making. So those are the types of connections and data sources that we're we're talking about. And, you know, just to give you some examples, engineering is definitely obvious, but we say OT operational data, we're talking about data coming from IOT feeds, sensors, cameras, etc. And informational or IT data, we're talking about data coming from asset registries, maintenance records and inspection records, and such.
Host: For a digital twin to work, does everyone have to be playing the game or can a team work if one member doesn't play?
Adam Klatzkin: So, sure it can work if one member doesn't play. I guess it depends on the team you're referring to. Are you referring to the engineering team? Which team do you have in mind?
Host: The project team, I guess would be, with all the different players that need to collaborate. To go back to our earlier example, to build a bridge, if one of the designers or one of the constructors decides not to play, can it still work?
Adam Klatzkin: It can still work. The value is only going to be great as the data that has been federated into the twin. So, if there are members that are keeping data out of it. There will be constraints on maximizing the value of all of the asset data in creating that digital representation of a physical thing you're trying to design and build. But it's absolutely still viable. And actually, our premise is to get started easily and quickly with the data you have at hand. There are use cases that can be developed out around it and you'll recognize more value as more data is connected into the system.
Host: Another thing that that you had in your documents, which was sort of an intriguing statement, was that opportunity and optimism are the value that a digital twin can bring to the infrastructure industry. I was struck by the word optimism as well as opportunity. What are you saying there?
Adam Klatzkin: We're seeing digital twins really present a new business opportunity for engineering services firms that have been exciting, exciting to us and exciting to them. And I say this anecdotally, from the many conversations I and our teams here have had with these accounts. We're talking to many that see it as a way to create new business opportunities and to deal with challenges such as an aging workforce, the labor shortage, shrinking profit margins, etc. There's an opportunity not just to do things more effectively and efficiently, but also an opportunity to create a new business model where they start delivering digital twin services to clients and therefore expand more into operations and maintenance, particularly from a data perspective. So to create that digital twin hand it over, but also maintain it into the future because it doesn't magically maintain itself.
Adam Klatzkin: There's effort involved in curating and continuing to grow the connections of information that may be accessible or pop up over time into this digital twin. So these firms are looking at the opportunity to kind of maximize the value of all of the owner's data into operations, provide digital leadership and transformation to the owner during operations. And they're really focused on following that data and maximizing the value of it over the entire asset life cycle. Pace pays dividends for the entire life of that asset. And the reality is somebody is going to do this. As I said, 33% of the penetration to date is being driven by owners. They're interested. So someone is going to do it. Who else is better situated than the engineering firms that have created the content. They understand the data best. It really distinguishes them from a run-of-the-mill systems integrator that's going to raise their hand and fill that void if nobody else is there to do it. So that is what is generating a lot of the opportunity and optimism around the value of digital twins beyond just the efficiencies that they'll introduce to existing processes in engineering.
Host: And I would think another benefit of that would be you're maintaining the relationship with that client after the project is over. And that maintaining that relationship opens you up for potential business opportunities with that client in the long term.
Adam Klatzkin: Absolutely. Absolutely. Yes.
Host: For firms looking to use a digital twin, Bentley recommends being flexible. What do you mean by being flexible in this context?
Adam Klatzkin: There are certainly multiple perspectives from which to look at flexibility. One is we have to recognize that digital twins are emerging technology. This is not a mature solution that's been in the market for years. This is something that is being developed at cloud speed at the moment. And that comes with the recognition that change is going to happen quickly. The services that Bentley is pushing out, this isn't an old school product that's going through monolithic releases once a year. We're pushing out new functionality on a regular basis, fixes and new functionality. So there has to be flexibility from an organization's perspective in how to keep on top and to learn and to increase their maturity with this technology as new capabilities are introduced. So flexibility to learn and pick up quickly what's coming out.
Adam Klatzkin: The second aspect I'd say for flexibility is in the use cases, scenarios and connections that you may establish and use. The digital twin is a Swiss army knife of sorts, so there are different things that you could consider and benefits that could be recognized and you should have flexibility in continuously evaluating those and your benefits over time. The third is, as I said earlier, we have an open-source platform for digital twins that we call iModel js. It's a development platform. Many engineering firms have their own in-house development teams, and if you do, so you should have the flexibility in those teams, evaluating the platform, leveraging how you might develop out automation integrations to bespoke systems or other systems if you have them, and new tooling you could add, even creating your own IP around the digital twin using this platform.
Adam Klatzkin: And then the last thing I'll say about flexibility is it's great to have a vision and a long-term plan, but don't focus on conquering the world from day one. You can get started quickly and easily with a digital twin on some scenarios and use cases that will bring immediate benefits. Find those lighthouse projects, get started, and understand the technology. And the long-term plan will certainly come out of that. But as this technology is emerging, things are going to change and you have to have the flexibility to adapt.
Host: In talking about this technology changing, you used the term "cloud speed." It's a rapidly changing technology. Where do you see it going in the next few years, beyond growing rapidly?
Adam Klatzkin: I'm most excited about is the potential, if I hone in on one, is the potential for machine learning. The amount of data that will be collected within the digital twin, due to all of the changes, everything that's being connected into the digital twin is going to make it very easy to take some things that may previously have required a lot of manual effort and to leverage artificial intelligence and machine learning in order to do those. Particularly I'm interested in the lessons learned that can be gleaned from a portfolio of projects and applying those to future projects. Identifying trends and patterns that resulted in mistakes or cost increases or safety incidents and leveraging machine learning to flag future projects as those types of scenarios come up again and to flag those in advance and provide early indicators.
Adam Klatzkin: But there is tons of opportunity for machine learning against this technology. That's certainly an area where I see this going in the next few years and it's going to add tremendous value. And the last thing I'll add there is, going back to the previous question on flexibility and where this is moving, it is incredibly important to get started quickly. Because there are many organizations that have already adopted. And if you have concerns and decide to let others lead the way and figure it out before you do, you're going to be left behind. And the potential business implications and benefits of changing business models around digital twins, if you miss the mark there, it could be catastrophic for your firm. So I really encourage everyone to start taking a look at digital twins, find that low hanging fruit, find those lighthouse projects and get started now.
Host: Well, I would agree with you after having listened to what these are and learning so much. It seems to me to be the future. I really appreciate your taking the time to speak with us. Thank you so much.