As a structural engineer with over 15 years in consulting firms and a parallel career exploring AI applications in design, I’ve spent countless late nights iterating on models, second-guessing load paths, and signing my name to drawings that carry real-world consequences. The question “Can AI design structures?” isn’t theoretical for me—it’s personal. It touches every phase of how we deliver safe, efficient, and economical buildings and bridges. The answer is nuanced: AI is already transforming how we work, but it is not (and will not soon be) a replacement for human engineers. Here’s a clear-eyed look at where we stand today, what’s coming, and how the profession must evolve. How Structural Design Actually Happens Today Structural design is not a linear calculation. It is an iterative, judgment-driven conversation between physics, codes, client needs, site constraints, and real-world constructability. The workflow typically unfolds like this: 1. Understand the structure’s purpose ...
Foundations fail in a number of ways. When we say a foundation fails in sliding , we are not talking about cracking concrete or soil punching. We are talking about something more fundamental: The entire foundation block tries to move sideways as a rigid body. Sliding vs Other Failures Bearing failure → soil beneath fails in compression (local or general shear). Overturning → foundation rotates about an edge due to moment. Sliding → horizontal forces overcome lateral resistance at the base. These are independent failure modes . A footing can be: Safe in bearing ✔ Safe in overturning ✔ But, Unsafe in sliding ✖ Why Sliding Is Often Ignored 1. Engineers focus heavily on vertical loads and soil bearing capacities. 2. Sliding does not “ look dramatic ” in drawings. 3. Friction is assumed to be “ automatically sufficient. ” There isn't heavy settlement if sliding failure occurs and you always have good frictional force that resists sliding which itself increases with vertical lo...