DR. ROSHAN KUMAR HOTA
Assistant Professor | Department of Mechanical Engineering
Indian Institute of Technology (IIT) Jodhpur, Rajasthan, India
Openings for Students
Ph.D. Positions Available
I am looking for highly motivated Ph.D. students to work on cutting-edge research in robotic grasping and manipulation.
- Strong background in Robotics, AI and Machine Learning, Controls, Computer Vision or related fields
- Experience with ROS2, Python, and MATLAB preferred
- Interest in Deep Learning and Reinforcement Learning applications
- Graduates in CS, Electrical, Electronics, Mechanical and other relevant areas are encouraged to apply.
M.Tech. Thesis Students
Multiple positions available for M.Tech. students interested in:
- Mechatronic design of robotic hands and grippers with tactile sensing
- Grasp planning and manipulation algorithms-- Reinforcement Learning and Foundational Models based approach
- Vision-based object recognition for grasping
- Integration of AI/ML techniques in robotics
- Kinematic analysis and motion planning of a symmteric space-robot
B.Tech. Projects/ Design credit projects
Undergraduate students can work on exciting projects including:
- Design and development of multi-fingered robotic hands - (reference: LEAP Hand/ ORCA Hand).
- Design and development of Flexible tactile sensors (reference 3D Vitac) and integration with robotic hands for manipulation tasks.
- Reinforcement Learning Based control.
- Computer vision for robotic applications.
- Integration LLMs for robotic applications.
- Simulation of robotic grasping and maanipulation (in Pybullet, Mujoco, Isaas Sim).
- Inference from Vision Language Action Model in using a simulated manipulator (Pybullet, Mujoco, Isac Sim).
- Design and development of a hand-held device for manipulation demonstration for foundational models in robotic manipulation.
Email at rkhota@iitj.ac.in to apply
About Me
I am Dr. Roshan Kumar Hota, an Assistant Professor at the Indian Institute of Technology (IIT) Jodhpur, with a Ph.D. in Mechanical Engineering from IIT Kharagpur. My main research interests are in the design of multi-fingered robotic hands, grasping and manipulation. I apply model-based and Reinforcement Learning based control methods for solving problems in robotic grasping and manipulation. This has applications in industrial automation and manipulation of objects in unstructured environments.
Research Highlights
Underactuated Hand Design
Five-fingered anthropomorphic hand actuated by a single motor for adaptive grasping
Vision-based Grasp Planning
Automated grasp planning using multiple quality metrics and deep learning
Flexible Gripping Solutions
Handling variety of objects including fruits and deformable items
Experimental Research
State-of-the-art robotic manipulation testbed with vision systems
Education
Thesis: Design Exploration of Underactuated Robotic Hands for Enveloping and Fingertip Grasps
Research Experience
Worked on the mechatronic design of multi-fingered robotic hands for manipulation research.
Developed flexible and robust gripping solutions for product variation. Collaborated with KUKA Robotics and Atlas Copco on grasp planning and gripper design.
Research Interests
Technical Skills
Selected Publications
Teaching
- Kinematics and Dynamics of Machines (MEL3020, Undergraduate 5th Semester)
- Advanced Robotic Manipulation (RML7120, M.Tech./ Ph.D.)
Thesis Supervision: M.Tech. and B.Tech. students in areas of robotic hand design, manipulation, and computer vision for grasping.
Patent Granted
Submersible pressure hull adapted for compressive forces in deep sea operations using additive manufacturing (Patent No. 444627)