Adaptive mechanics, learning and intelligent control improve soft robotic grasping

Handling soft, fragile, or slippery objects such as ripe fruit remains a challenge in robotics. Soft robotic graspers show tremendous promise in safely handling such objects without damaging them. Furthermore, creating software to control soft robots poses an additional challenge. In contrast, many animals with soft bodies solve this problem everyday as they forage and feed. Not only are they able to grasp and manipulate soft and fragile objects, but animals can also learn how to safely interact with new objects and vary how much force they apply during grasping based on their prior experience. This project will create a mechanism that can learn how to safely grasp a wide range of objects, including fragile foods like tomatoes and mushrooms. The ability for a robot to learn how to safely handle soft and fragile objects will have future applications in agriculture, manufacturing, and medicine.

Funding Agency: NSF - Foundational Research in Robotics Program 

Project Period: 2/2022 - 1/2025

Abstract Page: NSF-2138923

CMU Research Team

Dr. Victoria Webster-Wood

Victoria Webster-Wood

Faculty

Michael Bennington

Michael Bennington

Doctorate

Research interests
biomechanics, neuromechanics, biorobotics, and computational modeling
Kevin Dai

Kevin Dai

Doctorate

Research interests
bio-robots and controls
Ravesh Sukhnandan

Ravesh Sukhnandan

Doctorate

Research interests
soft robots, bioinspired robots, muscle mechanics
A generic human outline

Yu Wang

Masters

Research interests
bioinspired robotics, soft robotics, human-robot interaction

CWRU Research Team

Roger Quinn

Roger Quinn

Faculty

Hillel Chiel

Hillel Chiel

Faculty

Yanjun Li

Yanjun Li

Doctorate

Research Highlights

  • Viscoelastic McKibben actuators
    Tuning materials makes pneumatic muscles more biomimetic
  • Images of training results for a synthetic nervous system
    SNS discretization allows backprop and parameter learning
  • Images of training results for a synthetic nervous system
    SNS discretization allows backprop and parameter learning
  • An SNS controller architecture for a grasping robot
    SNS controllers can be developed inspired by animal nervous systems
  • Results from an SNS controlled gantry robot
    SNS control translates from sim-to-real with minimal tuning
  • Results from an SNS controlled soft grasping robot
    SNS control and tunable stiffness in soft grasping allows a range of objects to be pick-and-placed
  • Picture of PhD student training undergraduate
    This research is broadening participation in robotics through undergraduate research

    • Research videos

    Publications and Dissemination

    Peer-reviewed Publications

    * Indicate these authors contributed equally and may both list themselves as first author on this work.

    1. Y. Li*, R. Sukhnandan*, J. P. Gill, H. J. Chiel, V. Webster-Wood, R. D. Quinn. “A Bioinspired Synthetic Nervous System Controller for Pick-and-Place Manipulation”. ICRA 2023.
    2. M. J. Bennington*, T. Wang*, J. Yin, S. Bergbreiter, C. Majidi, V. A. Webster-Wood, “Design and Characterization of Viscoelastic McKibben Actuators with Tunable Force-Velocity Curves”. RoboSoft 2023.