Interested GT Undergraduate/Master Students?
Please fill out this Form.
Potential Topics for Thesis and Capstone Projects (updated on August 1, 2025):
- Compliant Loco-manipulation with Modular Soft Robots
- Description: We’re building a modular soft robot capable of both locomotion and manipulation in cluttered, uncertain environments. Projects span new module and connector hardware, controllers for compliant interactions, and learning-based policies (RL / VLA) for robust loco-manipulation. Students can focus on hardware, control, or learning, with shared sim-to-real evaluation.
- Desired skillset (not required for all, but some may apply): Mechatronics, CAD, embedded systems, Python/C++, control, reinforcement learning, MuJoCo/Isaac Gym, computer vision basics.
- Contact: Yilin Cai (yilincai@gatech.edu)
- Robotic Berry Harvesting with Soft Grippers
- Description: We aim to develop a gentle, reliable berry-picking system using soft grippers integrated with sensing and motion planning for field conditions. Work includes gripper/mechanism design, system integration (robot + vision + actuation), control/modeling for grasp and detachment, and benchmarking on real fruit. Subprojects fit both hardware-heavy and software-heavy interests.
- Desired skillset (not required for all, but some may apply): CAD, Mechatronics, Embedded systems, Python/C++/Matlab, Computer vision basics, Motion planning/control.
- Contact: Yilin Cai (yilincai@gatech.edu)
- Compact Mobile Platform and Soft Robotic Harvester
- Description: This project focuses on the compact packaging and integration of a mobile platform with a soft robotic arm for berry harvesting in real field conditions. The goal is to design a space-efficient, portable system that neatly houses actuators, sensors, power, and control hardware while maintaining durability and ease of deployment. Work includes CAD modeling, mechanical design, and integration of the soft manipulator with the mobile base, ensuring all components are optimally arranged for balance, accessibility, and field serviceability. Students can contribute to packaging-focused mechanical design, CAD layout, system integration, and prototyping of the complete platform.
- Desired skillset (not required for all, but some may apply): CAD (SolidWorks/Fusion 360), Mechanical packaging, Mechatronics, System integration, Soft robotics fabrication, Prototyping, Field-ready design.
- Contact: Milad Azizkhani (mazizkhani3@gatech.edu)
- Embedded AI for Medical Robotics in 3D Slicer
- Description: This project integrates a Large Language Model (LLM) directly into 3D Slicer, creating an embedded AI assistant that operates within the medical imaging and robotic control environment. The AI is context-aware, drawing on real-time scene data, imaging results, and robot state feedback. It enables natural language interaction, automated parameter selection, and intelligent command generation for the robot-control module.
- Desired skillset (not required for all, but some may apply): Python/C++, Artificial Intelligence, Imaging Processing, Mechatronics, Control
- Contact: Letian Ai (lai30@gatech.edu)
- Embedded Valves for Agile, Portable Soft Robots
- Description: Building on our success in achieving agile motion in soft robots, this project focuses on embedding miniature valves, pressure sensors, and control electronics directly inside soft pneumatic actuators. This integration eliminates bulky external valves and long pneumatic lines, allowing many actuators to be controlled from a single air supply and power source. The result is a sleek, portable architecture with faster response, higher actuator density, and the flexibility to implement advanced, intelligent control strategies. Students can contribute to actuator hardware design, embedded control systems, advanced control algorithms, or portable soft robot prototypes.
- Desired skillset (not required for all, but some may apply): Mechatronics, Embedded systems, Pneumatic actuation, CAD/3D printing, Control systems, Python/C++/MATLAB, ROS, Soft robotics design.
- Contact: Milad Azizkhani (mazizkhani3@gatech.edu)
- Physics Engine Development and Learning-Based Control for Soft Robots
- Description: We are developing a custom physics engine and communication protocols to accelerate research in learning-based control for soft robots. This platform supports the implementation of novel reinforcement learning algorithms, imitation learning strategies, and vision-language-action (VLA) policies, with a focus on sim-to-real transfer. Having achieved local simulation and control, the next step is extending our framework to real hardware for advanced, adaptive manipulation. Students can contribute to physics engine development, control pipeline design, learning algorithm implementation, and hardware integration.
- Desired skillset (not required for all, but some may apply): Python/C++, Reinforcement learning, Imitation learning, Vision-language models, Physics simulation, Control systems, ROS, Mechatronics, Soft robotics.
- Contact: Milad Azizkhani (mazizkhani3@gatech.edu)
- Adaptable Robotic Template for Needle Placement
- Description: Abdominal needle placement is common in oncology for procedures such as biopsies and ablations, but when practiced under MRI guidance, there are significant spatial and technical limitations. Continuing a lineage of robot development, we are working on a robotic version of a reusable, flexible needle guide. Work includes design of the robot as an extension of previous designs, fabrication, control, and benchtop evaluation.
- Desired skillset: CAD (specifically SolidWorks), mechatronics, MATLAB
- Contact: Sam Wilcox (swilcox33@gatech.edu)
- Patient Monitoring for Pressure Injury Prevention
- Description: You may struggle to believe that pressure injuries (PIs) are one of the largest costs to the US healthcare system every year. Part of the reason is because many factors go into PI formation and the healthcare works who help prevent them are unable to reliably track these factors while dealing with the primary conditions of their patients. We are working to develop a low-cost, bed-based system for tracking that indicators of PI with accompanying software for tracking intervention and predicting formation.
- Desired skillset: CAD (specifically SolidWorks), electronics, MATLAB, silicone casting, basic machine learning
- Contact: Sam Wilcox (swilcox33@gatech.edu)
- Physics simulation of continuum robot with full-body contacts in complex environments
- Description: This project aims to establish a simulation environment for continuum robot dynamics in cluttered workspaces, addressing contacts along the entire robot body. This will be either based on a suitable open-source physics engine, or in collaboration with topic 6 but with more focus on contact handling. If successful, the student can further extend their work to state estimation, control, or motion planning, based on their interest. All work will be validated on our continuum robot hardware. The immediate application of this project is MRI-guided robotic cardiac ablation, but it will also apply to general contact-rich tasks.
- Desired skillset (not required for all, but some may apply): confident in Python/C++ (required), robot kinematics and dynamics, ROS
- Contact: Yifan Wang (wangyf@gatech.edu)
- Surgical Autonomy based on da Vinci Research Kit (dVRK)
- Description: This project aims to develop advanced surgical autonomy capabilities for the da Vinci Research Kit (dVRK). Potential directions include integrating Imitation Learning for learning surgical skills from expert demonstrations, Long-Horizon Surgical Autonomy for planning and executing multi-step procedures, Tissue Interaction & Deformation Handling for safe manipulation of soft and deformable tissues, and Force Estimation from vision or kinematic signals to enable delicate, damage-free operations. Work will involve learning-based perception and control, hierarchical policy design, and sim-to-real transfer using platforms such as Isaac Lab and PyBullet.
- Desired skillset (not required for all, but some may apply): Experience with Python/C++, ROS, and simulation environments; Machine learning or imitation learning frameworks (PyTorch); Robotics control, Computer vision
- Contact: Yilin Cai (yilincai@gatech.edu)
- Contact Modeling and Control of Continuum Robots
- Description: Contact models have been extensively developed for rigid-link robots, but remain largely underexplored in the context of continuum robots (CRs). To address this gap, we aim to develop fast and general contact models that can enable reliable control of CRs. Our preliminary results indicate that complementarity constraints are well-suited for capturing both environmental contacts and internal friction in CRs. The student will contribute to integrating existing models, improving the computational speed, performing experimental validation, and developing novel controllers, e.g. model predictive control, for these models.
- Desired skillset: Matlab (required), C++, Mechatronics, fabrication.
- Contact: Jia Shen (jshen359@gatech.edu)
Prospective Graduate Students (we may have 1 PhD opening for 2026 Fall)
Ph.D. students: We can only admit students who have submitted the application to Robotics Ph.D. program. After you submit the official application, please read our most recent work to make sure you understand and are interested in my projects prior to your email to me. I receive a large number of template emails from prospective Ph.D. applicants. I do read your CV carefully, but due to time constraints, I can only respond to applicants who appear to be a strong fit for my lab’s focus areas. If you do not receive a reply, it means that your background does not align with our current work, and I encourage you to consider reaching out to other faculty members.
Requirements
- Self-motivation. Why do you want to do a PhD in our group? Is a PhD degree required for your career goal? How do you want to work towards your degree? You need to have a clear plan for your PhD study.
- Demonstrated solid research experience (for MS applicants) and working habits (for UG applicants). Have you worked in any research labs before? Is there a faculty member who can write a strong letter attesting to your research capabilities?
- Desired background in mechanical engineering, electrical engineering, automation, mechatronics, robotics, mechanics, EE, CS, etc.
- Experience with robotics hardware, software, and system development would be a plus.
- Excellent writing and communication skills.
- Excellent collaboration capability.
MS students: We typically don’t offer GRA positions for MS students. GTA position is subject to course requirement and teaching plan. We welcome MS students to perform research project in the lab via thesis or research hour format. Each master student will be supervised by a Ph.D. students. Please make sure you can find a PhD mentor if you decide to join the group. If you are GT MS student, please fill out this Form.