Can you share a bit about your background and journey to becoming an expert in automation, computational modelling, and structural engineering?
I attended Rutgers University in New Jersey, where I earned my degree in civil engineering. Afterwards, I worked for a civil engineering firm for two and a half years before deciding to return to college to pursue my master’s degree in structural engineering. I enrolled at Penn State University in the Civil Engineering Department, located in the State College, Pennsylvania. My master’s thesis focused on blast engineering, specifically designing containment facilities for the U.S. Navy, where they work on, build, and maintain missiles. It was an extensive master’s project that took two and a half years, and I graduated in 1995.
I began my career at Thornton Tomasetti in New York City as a junior engineer. A role that, while no longer in use, was what we called it at the time. Early on, I was heavily involved in technical engineering. Back in 1995, engineers didn’t commonly have personal computers, so I started my career doing mostly hand calculations and some computer work on a shared computer. Over the years, I progressed from technical work to leading projects. I left Thornton Tomasetti to work at a few different companies before returning in 2008.
Around that time, I developed an interest in computational modelling and automation. Computer programs and parametric modelling were still relatively new, but I was part of the Structural Engineers Association of New York, where I was a member of the programs committee. I co-organised an event on computational design, inviting experts in the field. This was around 2006 or 2007, and I was amazed to see what some people were already achieving with computational design. That experience helped me recognise the importance of leveraging this technology.
When I rejoined Thornton Tomasetti, the company became one of the first structural engineering firms to use building information modelling (BIM) for 100% of its new design projects. I was an early adopter, helping integrate 3D and BIM modelling into the company. I was also involved in external marketing around BIM and 3D modelling, as we were among the pioneers in this area, working under our CEO at the time, Thomas Scarangello. Initially, we had a small team of two or three people creating plugins and automation applications for various software. Around 2010, I took over and expanded the team’s capabilities, focusing not only on computational modelling but also on custom software development.
How do you see AI impacting the field of engineering, particularly in relation to steel structures? Can you provide examples of how AI and automation are currently being utilised?
The way I see AI is that, for me, it’s very similar to how we used tools in the past. As you know, AI requires extensive data to learn from, and once it has a robust dataset, training the model makes it incredibly effective. When it becomes accurate enough, AI can almost replicate the process we used in the pre-computer days, using tables and charts for design. Back then, we relied on precomputed designs in charts, like those found in steel or concrete manuals, to quickly look up results. In effect, AI is doing the same thing now, but on a massive scale, generating solutions for a wide variety of structural design elements — steel structures, trusses, brace frames, floor framing, columns, and even steel connections.
If we can train machine learning models to be highly accurate, we can rely on them, especially in the early design stages, when the design is still fluid and changing. This is where I see AI playing a significant role in steel structures.
In the recent model being developed by us, entire building components, from columns to slabs, floor framing, and brace frames, are now predicted by AI. Given the vast amount of data we’ve worked with, we can model structures in multi-material designs, covering steel, concrete, and timber. We can make design adjustments almost in real-time, within seconds, allowing us to rapidly redesign a building. Although it’s not 100 per cent accurate, it’s very close. This allows owners, clients, and architects to select a design they prefer quickly, enabling us to move on to more detailed aspects of the project.
This example is the culmination of years of work, and we believe it represents the future of how engineers will use machine learning and AI.
What do you see as the biggest challenges and opportunities in incorporating advanced computational modelling techniques in structural engineering projects? What role does steel play in it?
I think the biggest challenge is twofold. First, there’s the practical aspect of the investment required to achieve what we’ve done. Thornton Tomasetti is a forward-thinking firm, and we understand that investment is necessary to build custom machine learning and AI tools. The value of these tools, from now into the near and far future, is tremendous.
One challenge I faced when learning structural engineering was the time it took to become proficient. Because I started with manual calculations, the learning process was very slow. These tools, however, allow young engineers to learn much faster.
The second challenge, which applies to any new technology in our field, is encouraging people to break out of their established workflows and adopt new technology. I’ve had some recent success with face-to-face meetings with younger engineers, where we review applications like the ones I’ve mentioned. They understand that these tools make designs easier to navigate, which not only improves their work but also enhances their daily experience. By saving time, they can focus on more challenging, complex tasks instead of repeating the same design work.
What challenges and opportunities do you encounter when detailing architecturally exposed structural steel, and how do you ensure both aesthetic appeal and structural integrity?
I have worked on many architecturally exposed structural steel (AESS) projects, and the primary challenge is ensuring that contractors have the necessary skills and understanding of what AESS truly entails. Specifically, it has various quality levels, and it’s essential for contractors to understand what architects are looking for in terms of finishes, quality, and aesthetics.
This alignment on quality expectations is critical, which is why, on any AESS project, we typically require performance mock-ups. This means we ask contractors to build sample pieces representing the project, allowing us to visually assess the quality, finishes, and craftsmanship. These mock-ups usually set the baseline for the entire project.
From your experience, what differences do you observe in construction practices globally compared to those in India, particularly concerning steel structures?
My understanding of India’s steel practices is that it’s relatively less mature compared to countries like the US, where building with steel structures has a much longer history, dating back to the early 19th century. For instance, the building I’m in now was built in the early 1900s and is entirely a steel structure. This maturity reflects an advanced industry, from steel fabricators to steel fabrication itself, as well as how projects are delivered.
At Thornton Tomasetti, we actually have a division that handles steel fabrication engineering and complete building modelling, including connection modelling for large steel projects. We deliver a 3D model containing all the connections, bolts, welds—everything needed for fabrication. This approach largely replicates what steel fabricators have been doing here for many years.
In India, I think it’s similar to the situation with AESS: it depends on the contractors’ experience with steel structures and how they prefer to receive project information. Even after the design phase, there’s a long process involving connection design, shop drawings, fabrication drawings, quality control, and inspections, all of which are critical steps before the steel even reaches the job site.
How do you incorporate principles of sustainability and resilience into your design approach, particularly in the context of steel construction?
Sustainability is not new, but there is now a much more significant focus on sustainability, decarbonisation, and resilience. Unfortunately, resilience is related to factors like global warming, severe storms, and rising sea levels, all of which impact our work. When I started in the 90s, we didn’t have to consider these aspects nearly as much, but now they are central because we have to design low-carbon structures that can endure a minimum 50 year lifespan. The metrics related to flooding, hurricanes, and storms have amplified over the years.
In terms of sustainability, we’ve developed various internal tools and plugins to measure the carbon footprint of our projects. With structural steel, one of the biggest factors is actually the steel’s place of manufacture and the energy sources used. For example, steel from coal-fired plants has a different carbon impact than steel from plants using more sustainable energy sources. While I’m not an expert in all these factors, I know that the energy source for steel production has a significant effect on carbon emissions.
Another impactful change is that steel strength has significantly increased over the years. When I first started, 36 KSI steel was standard, but now we can access steel up to 80 KSI, which is more than double the strength. This allows us to reduce the amount of steel required since performance levels are now so much higher.
Finally, resilience ties into understanding future climate patterns and building in enough resilience. While this often involves architectural decisions, like placing buildings to avoid flood zones, structural engineers must also address resilience factors like wind and storm durability.
Can you share any ongoing projects that you’re particularly excited about, either due to their scale, complexity, or innovative aspects?
I’m not working on it directly, but our firm — my office here, along with my colleagues — is involved in a major project in Las Vegas, which is now public, so I can mention it. It’s a stadium in Las Vegas for a baseball team called the A’s, formerly the Oakland A’s, who have moved to Las Vegas. The design architect is the Bjarke Ingels Group, based out of New York and Copenhagen with offices around the world including India along with the Architect of Record HNTB, a US-based AE firm. This is an incredibly complex, large-scale project with a massive roof. It is expected to feature what will be the largest glass wall in the world. This is precisely the kind of project where computation, parametric design, and the latest technologies truly shine because of the scale and complexity involved.
