BRITE Pivot: Integrating Fractal Theory and AI for Bio-Inspired Fastening Technologies in Robotic Assembly

This Boosting Research Ideas for Transformative and Equitable Advances in Engineering (BRITE) Pivot award aims to establish a systematic methodology for bio-inspired design using artificial intelligence (AI) and mathematical sciences. This methodology will be applied to design probabilistic fasteners that facilitate robotic assembly operations. Robots face fundamental limitations in handling complex assembly tasks, a challenge that can be addressed by drawing inspiration from biological attachment mechanisms such as gecko feet, burrs, and bird feathers.

Despite the significance of bio-inspired design, no standardized methodology currently exists to help designers adequately learn from nature. This project advances AI and pattern recognition algorithms to systematically assess images of biological systems and detect visual characteristics such as shape, distribution, and density. Also, it utilizes fractal geometries to perform dimensional analysis of biological mechanisms.

Fractal geometry is a mathematical framework used to describe complex, irregular, and self-replicating structures, characterized by self-similarity, where smaller parts resemble the whole at varying scales. The resulting bio-inspired probabilistic mechanical fasteners address long-standing challenges in robotic assembly by providing secure, repeatable attachments without requiring expert precision or complex alignment.

This project promotes scientific progress through innovation at the intersection of engineering design, mathematics, robotics, and biology. Furthermore, it aligns with national interests by revolutionizing smart manufacturing and various industries reliant on robotic assembly, while advancing STEM education through undergraduate research experiences, webinars, workshops, and K-12 activities.

The objective of this project is to develop a bio-inspired design methodology powered by AI and fractal geometries with a focus on the design of probabilistic mechanical fasteners suitable for robotic assembly tasks. The objective is achieved through five main research tasks: (1) advancing pattern recognition algorithms to extract essential patterns within biological mechanisms and facilitate the conceptual design phase by identifying target and fractal-based patterns, (2) advancing dimensional analysis techniques based on fractal geometries to accelerate the detailed design process, (3) developing cutting-edge deep learning models for three dimensional reconstruction of biological attachment mechanisms, which integrates semantic information and employs noise reduction strategies, (4) systematic optimization and conducting of functional analyses to fine-tune mechanical fasteners parameters, and (5) performing simulation analysis for the generation of synthetic data, coupled with the real-world robotic assembly experiments to evaluate the performance of bio-inspired mechanical fastener prototypes.

The educational plan includes offering summer research experience for undergraduates, workshops and webinars, and training of K12, undergraduate, and graduate students with an emphasis on underrepresented groups.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.