http://robotics.usc.edu/~kale/hopper.html

1
Hopping Robot Project Principal Investigator
Gaurav S. Sukhatme Graduate Student Kale
Harbick Undergraduate Students Michael Poole and
Zach Turner
http//robotics.usc.edu/kale/hopper.html kale_at_rob
otics.usc.edu
Robotic Embedded Systems Lab
Robot Design

• Advantages
• On-board Power
• On-board Computation
• Self-contained
• Use of pneumatics allows changing of spring
  constants on the fly

• Disadvantages
• Short run-time
• Pneumatics more difficult to control than motors
• Air tanks are not as convenient as batteries

1. Kale Harbick and Gaurav S. Sukhatme. Height
   control for a one-legged hopping robot using a
   one-dimensional model. Technical Report
   IRIS-01-405, Institute for Robotics and
   Intelligent Systems, University of Southern
   California, 2001.
2. Kale Harbick and Gaurav S. Sukhatme. Height
   control for a one-legged hopping robot using a
   two-dimensional model. Technical Report
   IRIS-01-406, Institute for Robotics and
   Intelligent Systems, University of Southern
   California, 2001.
3. Kale Harbick and Gaurav S. Sukhatme. Controlling
   hopping height of a pneumatic monopod. ICRA 2002
   (to appear).
4. Kale Harbick and Gaurav S. Sukhatme. Speed
   Control of a Pneumatic Monopod using a Neural
   Network. Submitted to SAB 2002.

GUI screenshot
Height Control
Speed Control
Raibert
Neural Net
2d Model
1d Model

• Neutral point foot position which will result in
  zero net acceleration
• Raibert approach
• Approximated using stance duration
• Linearity assumption breaks down at higher speeds
• Neural network approach
• A simple 2-layer network can approximate the
  neutral point function with greater fidelity