• Matt.Home


Kod*lab Menu



Internal Links (Login Required)



Matthew Hale

Email: matthale [at] seas.upenn.edu

I am a Senior in the School of Engineering and Applied Science at the University of Pennsylvania. I am majoring in Electrical Engineering and minoring in Mathematics. I have worked in the KodLab since April 2009.

Research Interests

I am interested in problems of control where the controller has an incomplete model of the robot, only a basic model of the disturbances it will see (if it even has one), and/or limited capacity to sense its surroundings. The question I’m interested in is how to create behaviors robust to this uncertainty.

Past Work

In the past, I’ve done a lot of hands-on work with robots. I upgraded the old EduBot architecture to the new X-RHex Junior architecture, I revised a large portion of the electronics inside X-RHex, and I designed everything electronic that went into X-RHex Lite (XRL). I assembled the first XRL robot and delivered it to our colleagues at the Royal Veterinary College in London where they are using it for research into locomotion over compliant surfaces. A second XRL is in use at FSU for work on path planning, and a third is in the KodLab.

As part of my work with X-RHex, I traveled with several other Kodlab Members to Boston Dynamics in October of 2010 to demo the Smart Stand behavior on the X-RHex platform and traveled again in November of 2011 to the Mojave Desert to demo and develop hill-climbing behaviors on the X-RHex platform. The hill-climbing behavior we developed with uses an IMU to perform a simple gradient ascent behavior and a laser scanner to avoid obstacles as it climbs. It also has a yaw controller on it which minimizes the robot’s yaw to prevent gravity from causing a roll failure when the robot is too far out of alignment with the hill’s gradient.

Later, I worked on tuning a stairclimbing behavior for X-RHex. I tuned the actual climbing behavior by modifying the previously used setpoints in the robot’s open-loop controller used while it climbs. I also reworked the way in which the robot explores landings. This work resulted in a behavior that was very stable on a wide variety of stair geometries and which better identified stairs to climb and avoided falling down stairs as it explored landings. This and our hillclimbing work were the basis for our SSRR paper that was published in 2011 (info below).

Present Work

At present, I am working on a new way of measuring motor core temperatures on our robots. Our motors’ cores are made of copper and copper exhibits an affine relationship between electrical resistance and temperature. Using only current and voltage measurements, I am able to simply use Ohm’s Law to measure each motor’s electrical resistance and then calculate temperatures to within ±5C of their actual value. I am currently working on applying this temperature sensor to behaviors on X-RHex and am working towards publishing it soon.

Publications

  • Aaron M. Johnson, Matthew T. Hale, G.C. Haynes, and D.E. Koditschek, “Legged Hill and Stairwell Ascent,” SSRR 2011.
  • Matthew T. Hale, Aaron M. Johnson, and D.E. Koditschek, “Novel Motor Core Temperature Sensor for a Legged Robot,” in preparation
  • Simon Wilshin, G. Clark Haynes, Matthew T. Hale, Caragh Kelleher, and Andrew J. Spence. “Control strategies for legged locomotion on soft surfaces: insights from dogs and robots,” J. Roy. Soc. Interface. In preparation.

Copyright Kodlab, 2017