UROP Project

machine learning, flow control, flow experiments

Research Mentor: kshoele@fsu.edu Kourosh Shoele,
Department, College, Affiliation: Mechanical Engineering, FAMU-FSU College of Engineering
Contact Email: kshoele@fsu.edu
Research Assistant Supervisor (if different from mentor):
Research Assistant Supervisor Email:
Faculty Collaborators:
Faculty Collaborators Email:

Looking for Research Assistants: Yes
Number of Research Assistants: 2
Relevant Majors: Mechanical Engineering, mathematics
Project Location: On FSU Main Campus
Research Assistant Transportation Required:
Remote or In-person: Partially Remote
Approximate Weekly Hours: 8,
Roundtable Times and Zoom Link: Thursday September 5: 3-5

Project Description

Soft robotics are increasingly being used in underwater applications due to their flexibility, adaptability, and ability to navigate complex aquatic environments. Unlike traditional rigid robots, soft robots can deform and conform to their surroundings, enabling them to move efficiently through water, navigate tight spaces, and interact gently with delicate marine life or fragile underwater structures. These capabilities make soft robots ideal for environmental monitoring, marine biology research, underwater exploration, and search-and-rescue missions. Their soft, compliant materials allow safer interactions with the environment and living organisms, reducing the risk of damage or injury. Additionally, many soft robots are inspired by aquatic animals, making their movement patterns more energy-efficient and versatile in the dynamic underwater environment. As this technology advances, soft robots have the potential to expand the range of underwater operations and provide new tools for scientific discovery, conservation efforts, and industrial applications. The system to be useful needs to be carefully controlled, which for soft robots comes with extra complexity. Here, we explore how to use machine learning, computational model and towing tank experiments to identify the optimal way to improve the perform of flexible soft robots.

This study explores the application of reinforcement learning (RL) to optimize the control of tail motion in a soft swimming robot. By leveraging RL, the robots that are equipped with piezoelectric tails autonomously learn efficient tail movements that enhance propulsion and maneuverability in aquatic environments. The soft, flexible tail allows for a wide range of motion, mimicking natural swimming behaviors observed in aquatic organisms. Through iterative training, the RL algorithm adjusts the control policy to maximize swimming efficiency, stability, and agility, while adapting to varying fluid dynamics and external disturbances. this is done considering the limitations imposed by tradition force producing units in the robot and with the consideration of its flexibility. The expected results should demonstrate the potential of RL in advancing the performance of soft robots for underwater exploration, environmental monitoring, and bio-inspired robotics.

Research Tasks: -help in performing towing tank experiments
-reading literature
- postprocess the data
-learn how to measure forces and perform robotic experiments for underwater applications
-scientific presentation



Skills that research assistant(s) may need: Required: Creativity and interest in interdisciplinary research
Recommended: Desirable previous basic classes in fluid mechanics and/or robotic, and dynamics