Greg Bentley, CEO, Bentley Systems, sits down for a candid conversation about open source and how it helps the construction industry go digital. He explains that it is possible to have an open information environment with a digital twin, which can be used for analytics and machine learning. He adds that what has been the inhabitation in infrastructure going digital is the work of engineers is cloaked in obscure file formats.

Below is an excerpt from the interview. To hear the entire interview on The Peggy Smedley Show, log onto, and select 02/12/19 from the archives.

Peggy Smedley:
Why is open source valuable to the construction industry? Because we’ve seen an industry that many have said for years is laggard, but you and I both know that’s not true. This industry is forward thinking and that’s what’s made them so successful through many downturns.

Greg Bentley:
What we want to help with, we call it going digital, but we mean something in particular by going digital, Peggy. We say when the results of one computerized workflow can be the beginning of another computerized workflow and they can be linked together and automated and replicated, you then have digital workflows that can improve the quality and the throughput of what we do in construction.

And our realization is that the scope of benefit you can get from the repeatability and the automation of automating digital workflows. It’s another level of improvement you can get if you go beyond what you would connect within one discipline or one trade. And if you could have an open information project as a whole, for an infrastructure project delivery, the scope of automation from one workflow step to the next can be an order of magnitude more beneficial.

I think it wasn’t quite so possible as it now is to do this in a big way as we now can do with the what we call project digital twins.

How do you see that realization that you just described happening in a way that’s driving standards? I think we need to get much better information in infrastructure project delivery and to get closer so that we’re achieving all of these in an open information environment that we haven’t had, but the goal is to achieve what you just described. Is that really possible or are we still many years away from it?

I think it’s now possible to achieve this open information environment in the following way—if we galvanize around what would be called a project-digital twin. Now, what would you expect of a digital twin? Of course, since our construction projects are 3D, the environment needs to be 3D. Immersive visualization and augmented reality would be part of that. But there’s a breakthrough that makes that possible now. We call it reality modeling. On a site, you can capture the digital context and the technology we call reality modeling. It’s from overlapping photography or laser scanning, where you can bring in scans where necessary, but where the drones fly and the digital photographs overlap, we have software to resolve that into what we call a reality mesh.

It’s actually engineering ready and it represents the digital context of a site in a way that you can actually engineer to and relate to the BIM (building information modeling) models and then continuously survey that site, so that it’s an ongoing and synchronized digital twin.

So you need reality, which we get from reality modeling, the technology I described. For a digital twin though, you would want it explanatory capabilities. You’d want to be able to use the digital twin for analytics and machine learning. And that requires, if you like, aligning digitally, all of the engineering information of that project. And we can do that as well now, with the connected data environment and then synchronize the changes both from the engineering information side and the physical reality side.

And this is done now with a distributed database hub that synchronizes these changes and distributes them over a cloud service. And you actually can have an open information environment, which is accomplished through a digital twin and what everyone would expect from that. And that is only just feasible now. And it’s very interesting to have your podcast helping get the word out that we should have high expectations now for going digital in construction as a result.

With that digital twin, I can see the benefits on what’s happening in manufacturing. And you and I both know the benefits of the digital twin are more than just saying an overlay. You’re able to capture things that you didn’t even know before.

Peggy, you mentioned manufacturing. What we want to do is to industrialize, we could say, the process of project delivery. That is, to make it automated, to bring together offsite fabrication. The logistics, the interfaces, and so forth, are the stuff of going digital and digital twin to make this feasible. And it starts of course, with the design model and there always is a BIM model for an infrastructure project today…And now you mentioned AI and machine learning. The process of capturing the digital context, if you like, from the drones. We can now do machine learning to classify and recognize the digital components that relate to the BIM model and bring the cycle all together again, ultimately for asset management, when the project is delivered. So the digital twin and the digital twin concept pervading and surviving and improving throughout the whole project delivery process is what now, we can say, enables industrializing, replicating, and proving, automating.

In fact, on heavy civil projects, at the moment, we could say, if you like, for a new roadway project, the first autonomous vehicle on the roadway is the construction equipment because it’s guided from automated machine control that comes from the BIM models and the 4D construction modeling simulations. So it’s really possible to put this all together now and we just need to increase the impetus to take full advantage. I think the concept of digital twin provides that impetus.

How much will this change the way smart cities and our infrastructure? How are we building things compared to the way we did 10 years ago to how we’ll build things 10 years into the future?

The project digital twin should become the performance digital twin and will in the process that I’m describing, but there are so many ways to take advantage of this. What has been the inhibition in infrastructure to going digital is the notion that the work of engineers, which could be the digital DNA for the smart cities you’re describing, it’s been cloaked in obscure file formats that are only accessible to the engineer using the tool with which he created it.

So it’s not been possible to apply analytics and machine learning to the engineering information that drives our projects, until now. That digital twin is the way to do that…Putting that all together, we say opening up to that dark data is the seed change that will enable us to be thinking differently about how our work in construction and project delivery leads to smart cities and opportunities for engineers in improving the throughput and functionality of what’s designed in projects.

We believe at Bentley Systems, that our particular contribution to that should be opening up the dark data, the digital alignment, the change synchronization from the reality, the voracity of the engineering information and the integrity. Because if it’s going to be a twin, it has to continually change, as does a construction project or an infrastructure asset. But we want to be enabling everyone to take advantage of that…

How do you get or how can infrastructure projects get started with digital twins? I’m not sure that we’re seeing enough of them yet. Certainly not. We see more and more getting excited as we described in manufacturing. Are we seeing it at the infrastructure project level right now?

It’s integrated with the digital alignment of the EPC’s models and moving on to simulations and useful modeling, for instance, of flood resilience during construction and simulation…We call it an i-model.js, immersive environment analytics environment, is one where you can go on to and developers can download and GitHub, the immersive environment, to begin doing the augmented reality and the machine learning and AI analytics on construction sites, on construction data of theirs.

Will those become critical in order to be able to do things to make sure open source continues like Linux did? To make sure that they’re valuable? Do you see that that’s going to be something you’re going to follow down the road to keep going forward?

Well, considerations of open standards have been rather misunderstood in the infrastructure world. And we shouldn’t limit ourselves to open data standards, necessarily, in the world of going digital. If information can describe itself, it can be used for purposes beyond that for which it was created in these automated digital workflows.

But the open source communities are important. Where you have communities of contributors for new functional changes and then authorities of committers of those…

In today’s world of a bunch of nefarious characters, do you have to have someone policing an open-source environment?

Well, the very, most stringent use cases are now served with cloud computing and cloud computing is built on open-source software. The security layers are most critical and most important and are best done by the professionals who are providing these environments. We work very closely with Microsoft and Azure and our own secure environments. When we say opened up, of course, we mean data that was dark is opened up to secure access by authorized participants and industrial scale secure projects.

And I, myself, am an optimist. They’ll always be nefarious characters, but there are so many commercial incentives to improving security that gets better all the time, as well. And I think that individual engineering firms and infrastructure owners and so forth should, and generally do, acknowledge that their own environments are probably not and can’t be as secure as those that are professionally provided by the industrial scale players. And we’re glad to help with that.

Want to tweet about this article? Use hashtag #M2M #IoT #AI #artificialintelligence #machinelearning #bigdata #digitaltransformation #cybersecurity #blockchain #infrastructure #construction #technology #opensource #digitaltwin