AI-Driven Design and Engineering Applications

Integrating physics-based models, data, and intelligent optimization to design functional mechanical systems.

What do we do? 

MOA Lab develops computational strategies for discovering and optimizing mechanical structures with targeted dynamic functions. Our approach begins with physical understanding and extends it through numerical simulation, data-driven modeling, optimization, and artificial intelligence.

Architected materials involve strongly coupled design variables, including geometry, topology, material properties, contact conditions, internal resonances, and spatial distributions. As the number of variables increases, conventional parameter sweeps and trial-and-error design become inefficient.

We therefore seek to combine reduced-order mechanics, finite-element simulations, physics-informed models, and machine-learning algorithms. Physical constraints and known wave mechanisms are incorporated into the design process so that computational predictions remain interpretable and mechanically realizable.

Our research addresses both forward and inverse problems. Forward models predict how a given structure responds to dynamic excitation. Inverse-design frameworks begin with a desired function—such as a prescribed bandgap, localized mode, converted frequency, or impact response—and search for a structure capable of producing it.

Ultimately, we aim to establish a closed design loop connecting:

physical mechanism → computational model → intelligent design → fabrication → experimental validation

Why does it matter?

Advanced mechanical structures often contain too many interacting parameters to be designed efficiently by intuition alone. At the same time, purely data-driven methods may produce solutions that are difficult to interpret, fabricate, or generalize.

The AI combines the efficiency of machine learning with the reliability of mechanics. It can accelerate the exploration of large design spaces while preserving physical consistency and engineering feasibility. This framework can support the development of structures for vibration and noise control, impact protection, sensing, structural health monitoring, energy harvesting, adaptive materials, and intelligent mechanical systems.