Date of Award
Summer 8-7-2024
Degree Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
School
School of Computing
First Advisor
Iyad Kanj, PhD
Second Advisor
Isuru Godage, PhD
Third Advisor
Ljubomir Perkovic, PhD
Fourth Advisor
Umer Huzaifa, PhD
Fifth Advisor
Yue Chen, PhD
Abstract
Soft robot locomotion is a highly promising but under-researched subfield within the field of soft robotics. The compliant limbs and bodies of soft robots offer numerous benefits, including the ability to regulate impacts, tolerate falls, and navigate through tight spaces. These robots have the potential to be used for various applications, such as search and rescue, inspection, surveillance, and more. The state-of-the-art still faces many challenges, including limited degrees of freedom, a lack of diversity in gait trajectories, insufficient limb dexterity, limited payload capabilities, lack of control methods, etc. To address these challenges, this research introduces a modular approach to designing, modeling, validating, and controlling of soft mobile robots. The modular design philosophy aims at simplifying the robot construction and improving its reliability by focusing on designing and developing simpler soft robotic units. The research is conducted in two phases; (i) designing and fabricating modular soft mobile robots with different topologies and validating their fundamental locomotion gaits, and (ii) enhancing their locomotion capabilities through effective control strategies, i.e., closed-loop control.
During phase (i), inspired by spider monkeys' tails, a novel hybrid soft module is proposed. The soft module powered by pneumatics has an improved stiffness controllable range, and independent stiffness and shape control capabilities. As the first topology, soft modules are serially arranged to build soft robotic snakes (SRSs) that are wheelless, relying solely on spatial bending to achieve their movements. A kinematic model of the SRS is derived to achieve snake locomotion trajectories, namely sidewinding, serpentine, planar rolling, helical rolling, and curved surface locomotion. This is a significant improvement over the previous designs, which were either limited to planar movements or relied on wheels for locomotion. Additionally, a complete spatial dynamic model for the SRSs is proposed and experimentally validated. As the second topology, four soft modules are arranged in parallel to fabricate a soft-limbed robot that can mimic pinniped locomotion. A complete floating-base kinematic model of the proposed robot is derived to generate and experimentally validate a variety of locomotion gaits including a novel energy-efficient locomotion mode called tumbling. As the third topology, five soft modules are arranged serially and in parallel to fabricate a soft quadrupedal robot. Parameterized quadrupedal trajectories for crawling and trotting locomotion are derived utilizing the kinematic model of the robot. In crawling, gait models are derived to predict locomotion effectiveness, and experimental results confirm their prediction accuracy. A physics-enabled quadruped dynamic model is utilized to optimize and validate trotting locomotion trajectories. This modular approach provides a promising solution to the challenge of building high-dimensional soft robots capable of complex locomotion gaits and offers exciting possibilities for future research.
During phase (ii), closed-loop feedback control schemes are implemented to effectively manage the locomotion. Feedback control in mobile robots is important since it can track locomotion and perform dynamic locomotion adjustments necessary in real-world applications. Wireless sensors are integrated to measure the deformation of the SRS body and the quadruped limbs. The measured trajectory parameters are compared and adjusted in real-time using a Jacobian-based kinematic control system to match the intended locomotion trajectories. The results demonstrate that closed-loop controlled locomotion trajectories outperformed the previously tested open-loop control trajectories, significantly enhancing locomotion for field applications.
Recommended Citation
Kodippili Arachchige, Dimuthu, "Modular approach to soft mobile robots" (2024). College of Computing and Digital Media Dissertations. 60.
https://via.library.depaul.edu/cdm_etd/60