COMMITTEE CHAIR: Dr. Jaejong Park

TITLE: DESIGN AND INTEGRATION OF A BIOMIMETIC INTELLIGENT GRIPPER FOR OBJECT PERCEPTION AND GRASPING WITH THE FRANKA EMIKA PANDA ROBOT

ABSTRACT: Robots are becoming an essential part of our daily lives. Nowadays, they are widely used in factories, hospitals, and delivery services. However, one of the challenges for robot designers is creating a gripper that is flexible and adaptive to the objects it needs to hold, in order to provide more interactive capabilities with human beings. Most traditional grippers on the market rely on a single static setup, which makes it challenging to grasp new items and requires additional setup and programming. These grippers often fail because they cannot adjust their force or shape, which can lead to slippage or damage. To overcome these challenges, this work presents a scanner-based, underactuated robotic gripper that can adapt its shape and gripping force to the object it holds. The gripper is modified to be attached to a popular industry 7-degree-of-freedom robotic manipulator and combined with a structured-light scanner to perform a full 360-degree scan and create an accurate 3D model of the object. A computer program then calculates the most stable position and orientation for gripping. The gripper is inspired by the human hand for its high degree of freedom, which makes it more natural and flexible, and to make it adaptive to different shapes, much like a human hand does. Tactile sensing capability and an adaptive force control system of the gripper can fine-tune its gripping force to prevent slippage or damage to even delicate objects. 3D printing makes the design affordable, replicable, and easy to modify. Its customizable design and modular structure encourage students, researchers, and educators to experiment, add sensors, and improve the system. Additionally, integrating TinyML on these edge devices can enable the gripper to learn and adapt its grasping strategy more efficiently over time. This human-hand-like gripper could be developed into prosthetic hands for amputees and offer an affordable solution. This project combines a bio-inspired design and educational accessibility to create a flexible, low-cost, innovative robotic gripper suitable for industrial and learning applications, even prosthetic designs.

Keywords: Robotics, Biomimetic, Gripper, AI

Location Online:

Zoom Link: https://pvpanther.zoom.us/j/98954314331?pwd=Ws7m81eQtiRftXyC21n6SL60uTIXNC.1

Meeting ID: 989 5431 4331

Passcode: 853078