RI: Small: Collaborative Research: Evolutionary Approach to Optimal Morphology and Control of Transformable Soft Robots

The overall objective of the project is to leverage the principles of evolution in order to develop a new type of complex soft robots (robots that are constructed from compliant material) that can change their shape and mechanical properties to assume different forms in order to overcome environmental challenges such as uneven terrain. The motivation comes from the ways animals achieve efficient locomotion in their habitats by varying body shape and mechanical properties as a result of evolution over millions of years. Based on prior numerical results, the investigators hypothesize that highly adaptable locomotion can be achieved when given enough complexity in terms of shape deformation and mechanical properties. Current bioinspired soft robots, however, have limited capability to investigate such hypothesis. One key goal in this proposal is to develop innovative soft robotic platforms to study, for the first time, these numerical findings. The planned research has the potential to significantly enhance the functionality, adaptability, and versality of high-dimensional soft robots. The broader impact would be significant in several areas where soft robots have shown promise, particularly in inspection operations and search-and-rescue missions. The project is also expected to have a significant educational impact by engaging undergraduate students in the planned research activities and developing course content and new courses geared towards soft robots. Outreach efforts focus on workshops, public lectures, and engagements with elementary, middle, and high schools to foster collaborations with the local communities. The planned research is poised to unlock the evolutionary secrets of biological systems and open the door for the next generation of energy-efficient, adaptable, versatile, and human-friendly robots. To meet these goals, investigators propose to develop novel transformable and stiffness controllable soft robots and apply evolutionary computing to generate optimal morphological and stiffness trajectories for locomotion. The investigators employ evolutionary computing-based learning models to circumvent the curse of dimensionality associated with complex soft robotic systems and develop a comprehensive framework for developing optimal and robust control schemes and testing thereof using a series of rigorous experimental evaluations. 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.