By combing their fields of expertise, the researchers have developed a novel approach to design mechanically intelligent structures for offshore and underwater applications. Using AI, the researchers have been able to invert the design process. Normally, one has to do a multifold of iterations or wait for a “Eureka!” moment to come up with a design with tailored properties. In contrast, AI can help quickly and efficiently reverse engineer the optimal design directly starting from the desired property of a material or structure. Using this novel method, they have been building prototypes of novel structures for maritime and offshore applications, where extreme environments  impose atypical and case-specific design requirements for material and structural behaviors.
Subsequently, this novel design process is applied to smart materials giving access to functionalities that have not been achievable before. Now they can design novel metamaterials that have unique behaviors such as piezoelectricity-based energy harvesting capacity. Leveraging AI-driven acceleration, the researchers have been able to design and construct such exotic materials and structures from the micro to the macro scale in a much more efficient, creative, and faster manner.

One of the next steps for the researchers is to explore new and more complex material spaces: from piezoelectric to magnetomechanical and shape memory materials. On the AI side, the researchers are developing fundamentally new AI methods for mitigating the data-hungry nature of such AI-based design frameworks. Additionally, designing on the computer is one thing, but when the structure is deployed in real life – how can AI design human-centric and safe structures that are robust and resilient to the uncertainties of extreme environments?