What Is AI?

Nourbakhsh likes to compare AI—a branch of computer science—to a tree. Its roots are philosophy, logic, math, computation, cognitive science, biology and neuroscience. Its branches include computer vision, machine learning, speech recognition and synthesis, search and optimization, knowledge, representation, and reasoning and robotics.

AI scientists try to mimic human capabilities and design computer systems that are “artificially intelligent.”

“They can mimic how we see, how we talk, how we act, how we reason about objects and how we understand things,” explains Nourbakhsh.

AI In AEC

AI has recently been making headlines as part of very large technology companies (e.g., Alphabet, Meta, Apple, Amazon, Microsoft, etc.) making massive investments to train computers to analyze data and make human-like responses, even in literature, art and music. But how can AI fit in the specialized world of AEC?

In terms of the leaders and CEOs of the AEC industry, Nourbakhsh cites an example of a growing company that used an AI solution to augment sales capabilities without hiring a lot more people, efficiently expediting company growth.

For civil and structural engineers working in design, he suggests they map their processes and everyday work. “When you get a project, find out which areas energize you, and which tasks you don’t really like,” he explains.

From his personal experience working as a structural engineer, Nourbakhsh grew frustrated by having to start every project from scratch. So he looked into AI to train a machine-learning model that can examine past projects and come up with an initial solution that can work as a starting point. He found it was possible; as new projects come in, AI can propose a starting model so each project doesn’t have to start from scratch.

And with major staffing problems plaguing the industry, AI can help capture and preserve organizational knowledge.

“This technology can learn over time as more projects come in,” he notes. “After five or 10 years, this solution gets smarter and better, and it can help your future junior engineers do better work. It’s as if you have a senior structural or civil engineer in your company that won’t ever leave and can get better through time.”

The Other ‘Three Ps’

Through his career and research, Nourbakhsh identified “three Ps” that historically dominate the AEC industry and its projects: pressure, profitability and productivity.

Pressures vary for different personnel and projects, but for designers they include models that are getting more complex with shorter timeframes as well as increasing pressure from owners and clients who want more from the design.

On the construction side, he cites pressures caused by aging laborers and the lack of talent coming to the industry as well as increased demand from owners. “Sometimes we need to rebuild cities after hurricanes and earthquakes, and we need to be really agile and fast. This brings a lot of pressure.”

Nourbakhsh notes that lack of productivity, another “p,” is well known in the fragmented AEC industry because of regulations, labor shortages and slow intake of technology. And in terms of “profit,” he points out that the AEC industry is somewhat unique in that it often features high risks and low profit margins. Nourbakhsh believes AI can improve all the “three Ps.”

“Artificial intelligence is one of the most under-utilized emerging technologies in the AEC industry,” he says. “When understood and used correctly, it builds efficiency, empowers creativity and improves bottom-line results.”