PhD Defense: Redundant Design and Adaptive Control of an Interleaved Continuum Rigid Manipulator

This past Monday I completed my PhD in Mechanical Engineering, bringing to a close my years as a student and my time at UW-Madison.  While I expect to break out parts of my research and dissertation in greater detail in the coming weeks, for now I want to post my dissertation and a recording of my presentation. Many people have contributed to my experience at Wisconsin, so I thank my parents, my advisor Mike Zinn, my committee, and many friends for making these years enjoyable and fruitful.

My dissertation (.pdf, 105MB) and abstract:

Continuum manipulator compliance enables operation in delicate environments at the cost of challenging actuation and control. In the case of catheter ablation of atrial fibrillation, the compliance of the continuum backbone lends an inherent safety to the device. This inherent safety frustrates attempts at precise, accurate, and fast control, limiting these devices to simple, static positioning tasks. This dissertation develops Interleaved Continuum-Rigid Manipulation, by which the hysteretic nonlinearities encountered in tendon-actuated continuum manipulators are compensated by discrete rigid joints located between continuum sections. The rigid joints introduce actuation redundancy, which an interleaved controller may use to avoid continuum nonlinearities and dynamic excitations, or to prefer particular configurations that may improve task accuracy, permit greater end-effector forces, or avoid environment obstacles. Two experimental systems explore the potential of these joints to 1) correct for actuation nonlinearities and enhance manipulator performance and 2) increase the manipulator’s dexterous workspace.
These experiments expose important design and control observations that were not apparent in the general robotic and continuum literature.

My presentation: slides (.pdf, 8.4MB)

 

 

PhD Preliminary Examination

I gave my prelim presentation this past January, recapping my work with Prof. Zinn on Interleaved Continuum-Rigid Manipulation. The presentation went well and I enjoyed the audience’s and committee’s questions, most of which centered on things that would be very fun to look at given infinite time. As I hope to graduate soon, I’ll only be able to look into a few of the most fundamental questions, saving others for later students. With that, here are links to my document and narrated slides, followed by the prelim’s abstract:

Preliminary Thesis [45MB PDF]

Slides [6.1MB PDF]

Prelim Presentation from Ben Conrad on Vimeo. (and on YouTube)

Abstract

Continuum manipulator compliance enables operation in delicate environments at the cost of challenging actuation and control. In the case of catheter ablation of atrial fibrillation, the compliance of the continuum backbone lends an inherent safety to the device. This inherent safety frustrates attempts at precise, accurate, and fast control, limiting their use to simple, static positioning tasks. This preliminary work develops Interleaved Continuum-Rigid Manipulation, by which the hysteretic nonlinearities encountered in tendon-actuated continuum manipulators are compensated by discrete rigid joints located between continuum sections. The rigid joints introduce actuation redundancy, which an interleaved controller may use to avoid continuum nonlinearities and dynamic excitations, or to prefer particular configurations that may improve task accuracy, permit greater end-effector forces, or avoid environment obstacles. Two experimental systems explore the potential of these joints to 1) correct for actuation nonlinearities and enhance manipulator performance and 2) increase the manipulator’s dexterous workspace. These experiments also expose important design and control observations that were not apparent in the general robotic and continuum literature.

TRACLabs

After graduating in December 2010 I had 6 months to spend before starting grad school, so I went on an extended internship with TRACLabs in Houston, TX.  TRACLabs is a small company that does automation and robotics research and development work for NASA and the Defense Department.

My work centered on developing and using TRACBot, a powered wheelchair – based development robot.  During the spring and summer Patrick, Mars, Bryn, and I significantly upgraded TRACBot from

TRACBot outside
TRACBot outside

to

I worked on a number of projects over the spring and summer, here are some highlight videos:

Beat-Based Gestures with Brown University

Leader Following

Wounded Soldier Extraction

And some pictures:

ZeroG Team

I enjoyed developing three experiments with the ZeroG Team during my undergrad at UW-Madison.  Participating in NASA’s Reduced Gravity Student Flight Opportunities Program, I joined students from across the nation in designing experiments to be performed in a microgravity environment. Over three years, we submitted three research proposals, and NASA selected two of these experiments for investigation aboard the DC-9B Weightless Wonder.  This plane achieves microgravity (zero-g) by flying a parabolic path, during which the plane, occupants, and experiments experience zero gravity for 30 seconds.  My flights capped 11 consecutive years of UW experiments in this program; see below for descriptions.

Experiment Summaries:

2006-7: Measuring Capillary Forces in Microgravity

Our team designed an experiment to research capillary action in microgravity.  Capillary action is the phenomena that allows plants to transport liquid from their roots up to the highest branches.  In contrast to terrestrial applications, microgravity fluid systems cannot rely on gravity to collect fluids; instead the absence of gravity often ‘allows’ fluids to get trapped in the corners of tanks far from the tank exit.  In these cases, the fluid cannot be drawn out of the tank, and though unspent, is useless.  To better understand this situation, this experiment measured the flow velocities of two liquids as they flowed up five differently-angled surfaces (see video, below).  These velocities are a function of the fluid properties and the surface geometry; interpreting our results will guide fluid system design to avoid trapped-fluid scenarios.  This research can also be applied to transporting fluids in space without pumping.

Presentation

2007 – Capillary Action – John & Ben from Ben Conrad on Vimeo.

2007-8: Spray Cooling in Microgravity

In this experiment we researched the effectiveness of spray cooling in microgravity.  Spray cooling uses an array of nozzles  to create a turbulent fluid mixture on some hot device for the purpose of cooling that device.  The hot device may be a microprocessor or laser diode, where instead of a heat sink and/or fan (as you find in your computer), we’re using a liquid cooling loop with fluid sprays that impinge on the heated surface.  Whereas the amount of energy that heat sinks can dissipate limited by the velocity and ‘energy absorption capability’ of air, using a liquid provides a substantially greater capability.  A simple schematic of the test chamber is below, as are our documents.  This research was presented at the 2009 Space, Propulsion, and Energy Sciences International Forum in Huntsville, AL.

Spray Cooling overview

SPESIF Presentation

SPESIF Proceedings Paper

2008-9: Continued Investigation of Linear Spray Cooling

Our previous spray cooling experiment answered some questions but motivate a number of others.  In my Junior year we proposed a new array design and a more thorough characterization of spray cooling in microgravity.  The new array design was precisely machined from a single piece of aluminum, as opposed to the array of tubes used in previous linear spray arrays.  Unfortunately, NASA’s support was reduced by 2/3rds and we were not selected to fly this experiment.

The new array concept:

And result:

The machined array with visible spray holes.
The machined array with visible spray holes.

Previous Years:

As mentioned, UW has been active in this program for a number of years; here are some of the materials I’ve collected which document our research.

2003: REDUCED GRAVITY AEROGEL FORMATION:

The Effects of Gravity on the Structure and Chemistry of Mesoscopic Particles in Sol-Gel Systems and the Subsequent Effects on the Structural and Optical Properties of Derived Aerogels

2004: DIRECT SYNTHESIS OF AEROGEL FROMSUPERCRITICAL FLUID:

A New One-Step Process for Weightless Aerogel Production and a Novel Method for Supercritical Fluid Analysis

2005: Long Duration Fluorescence Imaging of the Richtmyer-Meshkov Fluid Instability

2006: Dynamic Fluid Flow Due to Capillary Forces in Microgravity