pr_old.html

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01//EN" "http://www.w3.org/TR/html4/strict.dtd">
<html><head>


<meta content="text/html; charset=ISO-8859-1" http-equiv="content-type"><title>EB1 Info for Pranav Bhounsule</title></head><body><big><big><big>Pranav A. Bhounsule</big></big></big><br>
<br>
I orginally made this page for letter writers. I added the green card timeline later.<br>
Just below the end of timeline is the information I sent the letter writers. <a href="pr.html#letter">Click here to go down quickly.</a><br>
<br>
<span style="font-weight: bold;">Green Card Timeline:</span><br>

2014<br>

Sep 8: Chair sign<br>

Sep 9: Dean&#8217;s signature and submitted to International office<br>

Oct 24: Sign from Vice Provost (International office mailed back)<br>

Oct 27: Spoke with Lawyer.<br>

Oct 28: Lawyer sent list of documents. <br>

<br>

2015<br>

Feb 15 (approx): Medical exam.<br>

Mar 17: UT approves payment<br>

May 6: Final signatures at the Lawyer&#8217;s office<br>

June 4: Sent to USCIS: I140, I-485, Adjustment of status, Advanced Parole, and EAD. Total = 10 months<br>

June 22: Notice from USCIS that I-485, I-765, and I-131 were received on 5 June and notice was created on 12 June. <br>

July 8: Biometrics at USCIS<br>

Aug 26: Arti I797C EAD and advance parole authorization received.<br>

Sep 4: Pranav I79C EAD and advance parole authorization received.<br>

Sep 15: Arti&#8217;s received EAD and advance parole authorization.<br>

Dec 1: Pranav received EAD and advance parole authorization.<br>

<br>

2016<br>

May 16 Attorney initiated Premium processing<br>

June 2 Arti received email that EAD processing has been sent to Potomac center.<br>

June 8 Send cheque to Attorney<br>

June 10 Attorney sends the premium processing<br>

June 14 Premium processing form received by USICS<br>

June 22 Email from USCIS that I-140 is approved<br>

Nov 14 I-485 RFE <br>

Nov 22 Did medical and filled forms and resent documents in response to RFE.<br>

Nov 29 Documents received by USCIS<br>

Dec 15 Send letter on financial loss to the company and signed by chair asking for quick response to RFE.<br>

<br>

2017<br>

Jan 6 USCIS approved green card. Status updated online. Letter and card yet to be received.<br>

Jan 10 Status updated to card mailed on USCIS website.<br>

Jan 11 Letter received through mail<br>
<span style="color: red;">
Jan 13 Received the green card through mail.</span><br>
<big><big><br>
<a name="letter"></a><br>
Applying for Permanent Residency under EB1 Outstanding Professors and Researcher Category (<a href="http://www.uscis.gov/working-united-states/permanent-workers/employment-based-immigration-first-preference-eb-1">info link</a>)<br>
<br>
<a href="sample_letter.docx">Template for download (DOCX) 0.5 MB</a><br>
<br>
<small><span style="font-weight: bold; text-decoration: underline; color: red;">The letter should confirm that the individual meets some of the following criteria:</span><br style="font-weight: bold; text-decoration: underline;">
1. Has received major prizes or awards for outstanding achievement in the field;<br>
2. Has authored books, or articles in professional journals;<br>
3. Has made original scientific of scholarly contributions to the field;<br>
4. Holds membership in associations that require outstanding achievements of their members, as judged by recognized experts;<br>
5. Has participated as a judge of the work of others, either on a panel or by serving as a reviewer for a journal;<br>
6. Has had materials published discussing the individual and his/her work</small><small><br>
</small><br>
<a href="cv_pbhounsule_long.pdf">Curriculum Vitae (pdf)</a></big></big> <br>
<br>
<big style="font-weight: bold;">
Experience:</big><br>

Assistant Professor, Dept. of Mechanical Engineering, University of
Texas San Antonio, TX, USA &nbsp;&nbsp;&nbsp;&nbsp; Aug. 2014 -<br>

Postdoctoral Researcher, Disney Research Pittsburgh, Pittsburgh, PA,
USA, &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp;
&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp;
&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp;
&nbsp;&nbsp;&nbsp;&nbsp; Jan.
2012 - July 2014<br>

Visiting Researcher, Robotics Institute, Carnegie Mellon University,
Pittsburgh, PA, USA, &nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp;
&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; &nbsp; Jan.
2012 - Dec. 2013<br>

<big style="font-weight: bold;"><big><br>
<small>Education:</small></big></big><br>Ph.D. (Mechanical), Cornell
University, Ithaca, New York,
USA,&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
Aug. 2006 - May 2012<br>

M.Tech. (Applied Mechanics), Indian Institute of Technology Madras,
Chennai, India, &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;
&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; Aug. 2004 - May 2006<br>
B.E. (Mechanical), Goa Engineering College, Goa,
India,&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
July&nbsp; 2000 - May 2004<br>
<br>
<br>
<big><big><span style="color: red;">Research Achievements </span>(<a href="pr.html#research">Summary &amp; Paper Info click here)</a>:<br>
</big></big><big><big><small><span style="font-weight: bold;"><br>
-- World Record (unbeaten since 2011): </span>Longest distance ever walked by
a legged robot on a single battery charge. 40 miles. (BigDog 12.6 miles) </small></big></big><big><big><small><span style="color: rgb(102, 0, 204); font-weight: bold;">(I developed the control algorithm and did extensive testing to make this possible)</span></small></big></big><br>

<br>

<big><big><small><span style="font-weight: bold;">
-- World Record (unbeaten since 2011):</span> Lowest Cost of Transport (power
per unit weight per unit speed) for a legged robot. 0.19 (Humans ~0.3,
ASIMO ~2, PETMAN ~5) </small></big></big><big><big><small><span style="color: rgb(102, 0, 204); font-weight: bold;">(I developed the model-based control design and implementation on the robot)</span></small></big></big><br>

<big><big><small>
</small></big></big><br>
<big><big>
<small style="font-weight: bold;">-- 5 Journal Papers.&nbsp; </small></big></big><big style="font-weight: bold;"><big><small><span style="color: rgb(255, 102, 0);">First author on all five papers.</span> </small></big></big><big><big><small><br>
&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; 1 Intl Journal of Robotics Research. (IJRR) Impact factor 2.503. <span style="color: rgb(102, 0, 204);">(</span><span style="font-weight: bold; color: rgb(102, 0, 204);">7
author paper. 100% results in this paper are from my work but was only
possible because other authors worked on various parts such as robot
building, state estimation etc.</span><span style="color: rgb(102, 0, 204);">)</span><br>
&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; 1 IEEE-Transactions on Robotics (TRO) Impact factor 2.65 <span style="color: rgb(102, 0, 204);">(</span><span style="font-weight: bold; color: rgb(102, 0, 204);">Single author paper</span><span style="color: rgb(102, 0, 204);">)</span><br>
&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; 1 Robotica Impact factor 0.894 <span style="color: rgb(102, 0, 204);">(</span><span style="font-weight: bold; color: rgb(102, 0, 204);">Single Author paper</span><span style="color: rgb(102, 0, 204);">)</span><br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1 ASME Journal of Dynamic Systems, Measurements and Control, Impact factor 0.76))<span style="font-weight: bold;"> <span style="color: rgb(102, 0, 204);">(90% work done by me)</span></span><br>
&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; 1 Dynamics of Continuous,
Discrete and Impulsive Systems Series B: Applications and Algorithms. </small></big></big><big style="color: rgb(102, 0, 204);"><big><small>(<span style="font-weight: bold;">Single Author paper</span>)</small></big></big><br>
<big><big><small>&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp; I have a low
citations. This is becuase all these papers came out in 2014 (this
year). Please avoid mentioning citations in the letter.<br>
<br><span style="font-weight: bold;">
-- 3 Conference papers (peer reviewed) </span></small></big></big><big style="font-weight: bold;"><big><small><span style="color: rgb(255, 102, 0);">First author on all three papers.</span></small></big></big><br>
<big><big><small>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1 Intl Conference on Humanoid Robots <span style="color: rgb(102, 0, 204); font-weight: bold;">(100 % work done by me)</span><br>
&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;  1 Intl Conference on Robotics and Automation </small></big></big><big style="color: rgb(102, 0, 204);"><big><small><span style="font-weight: bold;">(100 % workd done by me</span>)</small></big></big><br>
<big><big><small>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 1 Climbing and Walking Robots <span style="color: rgb(102, 0, 204); font-weight: bold;">(conference version of the IJRR paper above)</span><br>
<br><span style="font-weight: bold;">
-- Best Paper "Biological Inspired Robotics" at the Climbing and Walking Robots Conference 2012.</span><br>
<br><span style="font-weight: bold;"></span></small></big></big><big><big><small><br>
<big style="color: red;">Reviewing (26 times):</big></small></big></big><big><big><small><span style="text-decoration: underline;"><br>
Proposals<br>
</span></small></big></big><big><big><small>-- 1 time National Science Foundation (NSF) <br>
<br>
</small></big></big><big><big><small><span style="text-decoration: underline;">Journals</span><br>
</small></big></big><big><big><small>-- 1 time Transaction on Robotics (IEEE-TRO)<br>
-- 1 time Transaction on Mechatronics (IEEE-TMech)<br>
-- 1 time Robotica </small></big></big><br>
<big><big><small>-- 1 time Journal of Robotics and Computer Integrated Manufacturing. (RCIM) <br>
</small></big></big><big><big><small>-- 1 time Journal of Intelligent and Robotic Systems<br>
<br>
<span style="text-decoration: underline;">Conferences:</span><br>
</small></big></big><big><big><small>-- 8 times International Conference on Robotics and Automation (ICRA)<br>
-- 8 times International Conference on Robots and Systems (IROS)</small></big></big><br>
<big><big><small>-- 2 time Intl. Conference on Humanoid Robotics. </small></big></big><big><big><small><br>
-- 1 time&nbsp; Controls and Decisions Conference (CDC)<br>
-- 1 time International Symposium on Robotics and Mechatronics (ISRM)<br>
<br>
<br style="color: red;">
</small><span style="text-decoration: underline; color: red;">Media Coverage for 40 miles walking record</span></big></big><span style="font-weight: bold;"><br>
TV</span><br>

NTDTV - Cornell Students' Walking Robot Sets World Record 1 min 31 sec (<a href="http://english.ntdtv.com/ntdtv_en/news_northamerica/2011-05-13/cornell-students-walking-robot-sets-world-record.html">link</a>) (<a href="http://www.youtube.com/watch?v=_udRgSRoqdk">youtube</a>) (<a href="media/Ranger2011_NTDTV.mp4">youtube_cached</a>)<br>

Reuters - Walking robot sets record 1min 12 sec (<a href="http://www.reuters.com/video/2011/05/13/walking-robot-sets-record?videoId=210791677">link</a>)&nbsp;<span style="font-weight: bold;"><br></span><span style="font-weight: bold;"><br>Web<br></span>Dailymail - Marathon robot: 'Ranger' sets a world record by walking 40.5 miles non-stop on a single battery charge (<a href="http://www.dailymail.co.uk/sciencetech/article-1385897/Marathon-robot-Ranger-sets-world-record-walking-40-5-miles-single-battery-charge-stopping-touched.html">link</a>) (<a href="media/Ranger2011_dailymail.pdf">cached</a>)<span style="font-weight: bold;"><br></span>MSNBC - Robot walks 40.5 miles non-stop (<a href="http://cosmiclog.msnbc.msn.com/_news/2011/05/12/6632345-robot-walks-405-miles-non-stop">link</a>) (<a href="Ranger2011_msnbc.pdf">cached</a>)<span style="font-weight: bold;"><br></span>Engadget - Cornell's Ranger robot walks 40.5 miles on a single charge, doesn't even break a sweat (<a href="http://www.engadget.com/2011/05/11/cornells-ranger-robot-walks-40-5-miles-on-a-single-charge-does">link</a>) (<a href="media/Ranger2011_engadget.pdf">cached</a>)<span style="font-weight: bold;"><br></span>Cornell Chronicle - Robot walks a 40.5-mile ultramarathon without recharge (<a href="http://www.news.cornell.edu/stories/May11/rangerRobot.html">link</a>) (<a href="media/Ranger2011_CornellChronicle.pdf">cached</a>)<span style="font-weight: bold;"></span><br>

Gizmag - Ranger robot breaks its own endurance record (<a href="http://www.gizmag.com/ranger-robot-breaks-its-own-endurance-record/18628/">link</a>) (<a href="media/Ranger2011_gizmag.pdf">cached</a>)<span style="font-weight: bold;"><br></span>Popsci - Ranger robot sets a new distance record, walking 40 miles on a single charge (<a href="http://www.popsci.com/technology/article/2011-05/cornells-ranger-robot-sets-new-distance-record-walking-45-miles-single-charge">link</a>) (<a href="media/Ranger2011_popsci.pdf">cached</a>)<span style="font-weight: bold;"><br></span>Ubergizmo - Robot Ranger walks 40.5 miles on solitary battery charge, setting a new world record in the process (<a href="http://www.ubergizmo.com/2011/05/robot-ranger-walks-40-5-miles-on-solitary-battery-charge-setting-a-new-world-record-in-the-process/">link</a>) (<a href="media/Ranger2011_ubergizmo.pdf">cached</a>)<span style="font-weight: bold;"></span><span style="font-weight: bold;"><br></span>Tecca - Ranger robot trots 40 miles straight on 5 cents worth of electricity (<a href="http://www.tecca.com/news/2011/05/12/cornell-ranger-robot/">link</a>) (<a href="media/Ranger2011_tecca.pdf">cached</a>)<br>
<br>
<br>
<big><big><big><span style="color: red;">More information about my papers:</span><a name="research"></a><br>
</big></big></big><br>
<big><big>Journal Paper 1: Intl Journal of Robotics Research</big><br>
<br>
</big>P. A. Bhounsule,
J.
Cortell, A.Grewal, B. Hendriksen, J.G.D. Karssen,
C. Paul, A. Ruina. <a href="rangerIJRR2013.pdf">Low-bandwidth
reflex-based control for lower power
walking: 65 km on a single battery charge</a>. International Journal of
Robotics Research,&nbsp; vol.33 no.10, 1305-1321, 2014. <a href="rangerIJRR2013supplement.pdf">Extended
appendix.</a> <a href="http://ruina.tam.cornell.edu/research/topics/locomotion_and_robotics/ranger/ranger_paper/index.html">Link
to webpage.</a> <br>

<a style="color: red;" href="https://www.youtube.com/watch?v=KLepY1AsaRk"><span style="font-weight: bold;">(click for youtube video)</span></a><br>

<br><span style="font-weight: bold;">Intellectual Merit:<br>
</span>a)&nbsp; Develepment of&nbsp; low information (minimal gains and minimal sensing) control algorithms for walking robot.<br>
b) Demonstrated 40 miles non-stop walking on a single battery charge (a world record)<br>
c) Demonstarted lowest Cost Of Transport ever achieved by a legged robot. TCOT = 0.19 (a world record)<br>
<span style="font-weight: bold;"><br>
Abstract:</span><br>

No legged walking robot yet approaches the high reliability and the low
power usage of a walking person, even on flat ground. Here we describe
a simple robot which makes a small progress towards that goal. Ranger
is a knee-less 4-legged &#8216;bipedal&#8217; robot which is energeti- cally and
computationally autonomous, except for radio controlled steering.
Ranger walked 65.2 km in 186,076 steps in about 31 hours without being
touched by a human with a total cost of transport [TCOT &#8801; P/mgv] of
0.28, similar to human&#8217;s TCOT of &#8776; 0.3. The high reliability and low
energy use were achieved by: 1) development of an accurate bench-test-
based simulation; 2) development of an intuitively tuned nominal
trajectory based on simple locomotion models; and 3) offline design of
a simple reflex-based (that is, event-driven dis- crete feed-forward)
stabilizing controller. Further, once we replaced the intuitively tuned
nominal trajectory with a trajectory found from a numerical
optimization, but still using event-based control, we could further
reduce the TCOT to 0.19. At TCOT = 0.19, the robot&#8217;s total power of
11.5W is used by sensors, processors and communications (45%), motor
dissipation (&#8776;34%) and positive mechanical work (&#8776;21%). Ranger&#8217;s
reliability and low energy use suggests that simplified implementation
of offline trajectory optimization, stabilized by a low-bandwidth
reflex-based controller, might lead to energy-effective reliable
walking of more complex robots.<br>

<br>

<br>

<br>

<br>

<big><big>Journal Paper 2: IEEE-Transaction on Robotics<br>
</big></big>P. A. Bhounsule. <a href="simplest_limits.pdf">Foot placement in the
simplest slope walker reveals a wide range of walking solutions</a>.
Transactions on Robotics, Vol 30, Issue 5, June 2014.\<br>

<span style="font-weight: bold;"></span><br>
<br>

Abstract:<br>

We show that the simplest slope walker can walk over wide combinations
of step lengths and step velocities at a given ramp slope by proper
choice of foot placement. We are able to find walking solutions up to
slope of 15.42 degrees, beyond which the ground reaction force on the
stance leg goes to zero, implying a flight phase. We also show that the
simplest walker can walk at human sized step length and step velocity
at a slope of 6.62 degrees. The central idea behind control using foot
placement is to balance the potential energy gained during descent with
the energy lost during collision at foot-strike. Finally, we give some
suggestions on how the ideas from foot placement control and energy
balance can be extended to realize walking motions on practical legged
systems.<small><br>
</small><br>
<br>
<br>

<br>

<big><big>Journal Paper 3: Robotica</big></big><br>

P. A. Bhounsule. <a href="walking_model.pdf">Control of a compass gait
walker based on energy
regulation using ankle push-off and foot placement.</a>
Robotica, pp. 1--11, June 2014.<br>

<br>
<big><big><small>Abstract:<br>
</small></big><small>In this paper, we present a theoretical study on
the control of a compass gait walker using energy regulation between
steps. We use a return map to relate the mid-stance robot kinetic
energy between steps with the two control inputs, namely, foot
placement and ankle push-off. We show that by regulating the robot
kinetic energy between steps using the two control inputs, we are able
to: 1) generate a wide range of walking speeds and stride lengths,
including average human walking; 2) cancel the effect of external
disturbance fully in a single step (dead-beat control); and, 3) switch
from one periodic gait to another in a single step. We hope that
insights from this control methodology can help develop robust
controllers for practical bipedal robots.</small></big><big><big><small><small><br>
</small></small><br>
<br>
<br>
Journal Paper 4: ASME Journal of Dynamics, Measurements, and Control</big></big><br>

P.
A. Bhounsule, A. Ruina, G Steissberg. <a href="event-based-intermittent.pdf">Discrete Decision Continuous Actuation
control: balance of an inverted pendulum and pumping a pendulum swing</a>. Accepted ASME Journal of Dynamics Systems, Measurement and
Control&nbsp; <a style="color: red;" href="https://www.youtube.com/watch?v=GCGeDHKNzm4"><span style="font-weight: bold;">click for youtube video</span></a><br>

<br>

<span style="font-weight: bold;">Intellectual Merit:</span><br>
- Dead beat (perfect correction) of disturbances in finite time is
possible using a discrete controllers. Further, demonstrated that it is
possible to control systems with significant time delays. For example,
we balance a pendulum with a time constant of 0.33 sec by sensing
*once* per second. In other words, time delay was three times the
characteristic time constant of the system. The video above shows
demonstration on a simple pendulum. <br>
<br>
Abstract:<br>

In some practical control problems of essentially-continuous systems,
the goal is not to tightly track a trajectory in state space, but only
some aspects of the state at various points along the trajectory, and
possibly only loosely. Here we show examples in which classical
discrete-control approaches can provide simple, low input and low
output bandwidth con- trol of such systems. The sensing occurs at
discrete state- or time-based events. Based on the state at the event,
we set a small set of control parameters. These parameters prescribe
features, e.g. amplitudes, of open-loop commands that, as- suming
perfect modeling, force the system to, or towards, goal points in the
trajectory. Using this discrete decision con- tinuous actuation (DDCA)
control approach, we demonstrate stabilization of two examples: 1)
linear dead-beat control of a time delayed linearized inverted
pendulum; and 2) pump- ing of a hanging pendulum. Advantages of this
approach include: It is computationally cheap compared to real-time
control or online optimization; it can handle long time de- lays; it
can fully correct disturbances in finite time (dead- beat control); it
can be simple, using few control gains and set points and limited
sensing; and it is low bandwidth for both sensing and actuator
commands. We have found the approach useful for control of robotic
walking.<br>
<br>
<br>
<br>
<big><big>
Journal Paper 5: </big></big><br>
P.
A. Bhounsule. <a href="pdw_benchmark.pdf">Numerical accuracy of two
benchmark models of walking:
the rimless wheel and the simplest walker.</a> Dynamics of
Continuous, Discrete and Impulsive Systems Series B: Applications and
Algorithms, vol. 21, pp 137--148, 2014.<br>
<br>
Abstract:<br>
The eigenvalues of the Jacobian of the return map are used to quantify
the stability of discrete dynamical systems, such as, the rimless wheel
and the simplest walker. The accuracy this Jacobian, usually obtained
by finite differencing, depends on the step size. Even with the most
optimal step size, only moderate accuracy is obtained. Here, we obtain
the Jacobian by numerically integrating the gradient of the equations
of motion. For the rimless wheel, our eigenvalue estimate is accurate
to 12 significant digits and is better than 9 significant digits
obtained by finite differencing with optimal step size obtained by
Coleman [Dynamics of Continuous, Discrete and Impulsive Systems Series
B, 16, 2009]. We first show that our method is able to produce the
eigenvalues accurate to 12 signifi- cant digits obtained by known
analytical solution for the rimless wheel. This benchmark calculation
then permits us to make the claim that the eigenvalues of the simplest
walker, for which the analytical solution is unknown, obtained using
our method are accurate to 12 significant digits.<br>

<br>

<br>
<br>

<big><big>Conference Paper 1: Intl Conference on Robotics and Automation (ICRA)</big></big><a style="color: red;" href="http://youtu.be/KzdAme1xg_M"><span style="font-weight: bold;"><big><big> </big></big></span></a><br>

P.
A. Bhounsule, K.Yamane. <a href="sky_task.pdf"><span style="text-decoration: underline;">Iterative Learning of Inverse Kinematics with Applications to Humanoid Robots</span></a>.
International Conference on Robotics and Automation 2015 (submitted)&nbsp;<big><big> </big></big><br>

<a style="color: red;" href="http://youtu.be/KzdAme1xg_M"><span style="font-weight: bold;">click for youtube video</span></a><br>

<br>

Abstract:<br>

We present an iterative learning framework for accurate end-effector
tracking of kinematically redundant robots. The iterative learning
control update rule is in the task space and consists of adding a
correction to the desired pose of the end-effector based on tracking
errors. The corrected end- effector desired pose is then fed to an
inverse kinematics solver that computes a new joint position command
that is implemented using a joint level servo on the robot. Our
framework has the following benefits: (1) It involves very little gain
tuning. (2) It converges in a few trials. (3) It is able to find
solutions that are within joint limits. (4) It is robust to modeling
errors. (5) It works with the existing inverse kinematics solver, hence
re-programming the solver is not required. We validate our method by
testing it on a humanoid robot doing two tasks; drawing the figure
eight and serving a glass of drink.<br>

<br>

<br>
<br>

<big><big>Conference Paper 2: Intl Conference on Humanoid Robots<br>
</big></big>P.
A. Bhounsule, K.Yamane. <a href="ilc_animatronics.pdf">Iterative
Learning Control for High-Fidelity
Tracking of Fast Motions on Entertainment Humanoid Robots</a>.
International Conference on Humanoid Robots, 2013<big><big> </big></big><br>

<a style="color: red;" href="https://www.youtube.com/watch?v=HBDpIxyxeQI"><span style="font-weight: bold;">click for youtube video</span></a><br>

<br>

Abstract: <br>

Creating animations for entertainment humanoid robots is a time
consuming process because of high aesthetic quality requirements as
well as poor tracking performance due to small actuators used in order
to realize human size. Once deployed, such robots are also expected to
work for years with minimum downtime for maintenance. In this pa- per,
we demonstrate a successful application of an iterative learning
control algorithm to automate the process of fine tuning choreographed
human-speed motions on a 37 degree- of-freedom humanoid robot. By using
good initial feed-forward commands generated by
experimentally-identified joint models, the learning algorithm
converges in about 9 iterations and achieves almost the same fidelity
as the manually fine tuned motion.<br>
<br>
<span style="font-weight: bold;"></span><br>

<br>
<br>


</body></html>