2026-05-20T22:30:23Zhttps://keep.lib.asu.edu/oai/request

oai:keep.lib.asu.edu:node-2008892025-05-09T20:56:10Zoai_pmh:alloai_pmh:repo_items

200889 https://hdl.handle.net/2286/R.2.N.200889 http://rightsstatements.org/vocab/InC/1.0/ http://creativecommons.org/licenses/by-nc-sa/4.0 2025-05 128 pages Kodithyala, Raj Zhang, Wenlong Aukes, Daniel Barrett, The Honors College School of Manufacturing Systems and Networks Tech Entrepreneurship & Mgmt Engineering Programs Soft robotic manipulators, utilizing compliant materials such as fabrics and elastomers, provide significant advantages over traditional rigid robots, particularly in terms of safety, flexibility, and bio-adaptability. However, accurate modeling of these systems remains a challenge due to their inherently nonlinear behavior. This honors thesis presents a hardware system that reliably collects data from a pneumatically actuated soft robotic arm and a PINN model to predict future states of the soft robotic arm. The data collected serve as input for a physics-informed neural network (PINN), a machine-learning model that incorporates physical principles to predict the motion of the soft manipulator. Initial results indicate promising performance for state-dependent predictions. Future research will explore integrating more sophisticated physical models into the PINN to further enhance the accuracy and reliability of soft robotic system modeling. Soft Robot Neural Network Robot Model Hardware Development and Physics-Informed Neural Network Model of Soft Robotic Arm