Imagine, for a moment, that we’re stepping back five years into the past. Digital twin was just a buzzword. BIM (building information modeling), while discussed for decades, was only just beginning to gain any meaningful traction. Now, here we are today, witnessing an extraordinary rise in all these technologies—and what’s even more exciting is all this momentum will certainly drive new applications and transform business models unlike anything we have ever witnessed before. Now we can confidently say we find ourselves at a thrilling crossroads of technological growth poised to shape the next five years with promising unprecedented transformation.
The numbers tell the story. The growth for digital twins is huge—$293 billion by 2035, up from $20 billion in 2024. BIM adoption will be $22 billion by 2032, up from $8 billion. Geospatial analytics will be $174 billion by 2032, all forecasted by the various analysts’ firms.
“In the next five years, I think there will be a whole new generation and when I say new generation—and when I talk about generation I don’t talk about their birthday, I talk about their mindset—there will be a whole new generation of engineers that will be working on infrastructure projects that will have a team of AI (artificial intelligence) agents that will be helping them all day long to do a lot of the work in a better way, in a more informed way, in a more insightful way,” says Julien Moutte, chief technology officer, Bentley Systems.
He adds digital models will become the design deliverable and AI will leverage them extensively—but it could also go beyond this as well.
To better understand where we are headed, let’s take a closer look at where we are today with AI in infrastructure projects. Moutte says there are two ways you can use AI for infrastructure modeling.
Explicit modeling: The AI tools are doing the geometry, they are making the calculations, and the engineers are guiding those tools. Here AI will help the engineers automate a lot of the work that is applied.
Implicit modeling: The user creates a prompt, and the AI will generate and materialize content. For infrastructure, there may be a time when AI is able to materialize buildings and models of infrastructure designs from the void.
Moutte adds, “We will only trust those models if we have been able to validate those models with the structural tools that we are applying to the explicit modeling to make sure those models are trustworthy and that we can actually put them in the hands of users and citizens without putting them in danger. That is how we believe AI hallucinations need to be approached.”
He adds we should not be afraid of implicit modeling and that it needs to be grounded with the tools that we trust so that we can make sure we adhere to the code and the local rules, so those designs are safe.
At the end of the day, we need to trust the AI. As Moutte says, “Any data you use for AI needs to be good quality data. It needs to be trustworthy.”
Moutte believes for trustworthy AI to become a reality, we need to bring that unique combination of AI innovation together with very well established and proven engines that are trusted in the infrastructure sector.
Once we can, this will spur digital twins and modeling into a whole new era of work. “What is really important is the engineers when they need to make decisions about the project they are working on, they need to have as much context as possible,” says Moutte.
The future of digital twin, AI, and BIM are certainly bright. It will be interesting to look back in five years and see how far we have come.
Want to tweet about this article? Use hashtags #construction #IoT #sustainability #AI #5G #cloud #edge #futureofwork #infrastructure #digitaltwin #design
