Purdue University
Mechanical Engineering

Vine Robots are a unique type of soft, continuum robot. Unlike traditional robots, which move due to surface contact, vine robots rely on growth for movement, similar to vines and plant roots. This growth happens via eversion, which adds material at its tip from the inside due to a provided internal pressure. This locomotion enables unique methods for environmental exploration and haptic interfaces.

To advance environmental exploration, this project aims to develope a vine robot with an end cap to deploy insect-scale robots in unstructured, dynamic environments for Non-Destructive Evaluation (NDE) and sensing. The end cap houses an IMU to sense the robot’s orientation and multiple cameras to explore the environment. In combination with a physics-based simulation, the end-cap design enables autonomous control and motion planning in contact-rich environments.

This project aims to make soft robots easier and more effective to use by developing shared-control strategies for real-world tasks. Soft robots are naturally safe and flexible, and by combining this with a human operator’s intuition about environmental affordances, we seek to enable emergent behaviors and expand the robot’s operational capabilities through real-time measures of human workload and robot uncertainty.

In collaboration with Dr. Denny Yu’s group at Purdue University

During kinesthetic teaching, robots gain information through physical interaction when humans kinesthetically guide it throughout the task. While prior works focus on how the robot learns, we are developing haptic systems that provide feedback to the human teacher. Our soft haptic displays wrap around and conform to the surface of a robot arm, adding haptic feedback at existing points of contact without significantly affecting the interaction.

In collaboration with Dr. Dylan Losey's group from Virginia Tech.

By leveraging the complementary capabilities of soft vine robots and rigid centimeter scale RESCUE rollers, this project enhances their combined performance while addressing the limitations each system faces individually.  Collaboration includes physical interactions in addition to transfer of both data and electric.

In collaboration with both Dr. Zeynep Temel and Dr. Melisa Martinez's groups from Carnegie Melon University

Soft robots are notoriously difficult to model due to near infinite degrees of freedom. Through various geometric and physics inspired approaches, this project aims to reduce a vine robot's kinematics to a more manageable model and then leverage this model to predict environmental interactions during navigation of environments.

In collaboration with Dr. Zachary Kingston's group from Purdue University

This project aims to develop easy-to-manufacture pneumatic actuator muscle with tunable force-to-strain relation. By adjusting the geometry of the actuator, the contraction and force produced by the actuator is physically programmed, which enables sophisticated patterns of motion with simplified control. The device can be used to actuate a wide range of devices, from haptic devices to continuum robots.