UTCS Colloquia/AI - Geoff Hollinger/University of Southern California, "Robotic Decision Making for Sensing in the Natural World", RLM 6.116

Contact Name: 
Jenna Whitney
Date: 
Nov 18, 2011 11:00am - 12:00pm

There is a sign-up schedule for this event that can be found at

http://apps.cs.utexas.edu/talkschedules/cgi/list_events.cgi

Type o

f Talk: UTCS Colloquia/AI

Speaker/Affiliation: Geoff Hollinger/Univers

ity of Southern California

Talk Audience: UTCS Faculty, Undergraduate
Students, Graduate Students and Outside Interested Parties

Date/Time

: Friday, November 18, 2011, 11:00 a.m.

Location: RLM 6.116

Ho

st: Peter Stone

Talk Title: Robotic Decision Making for Sensing in the
Natural World

Talk Abstract:
There is growing interest in the us

e of robots to gather information from natural environments. Examples inclu

de biological monitoring, mine sweeping, oil spill cleanup, and seismic

event detection. The increasing capabilities of the robots themselves enabl

e more sophisticated decision making techniques that optimize information g

athered and adapt as new information is received. The question becomes: how
do we develop path planning algorithms for information gathering tasks tha

t are capable of dealing with the communication limitations, noisy sensing

, and mobility restrictions present in natural environments? This talk con

siders two problems related to path planning for Autonomous Underwater Vehi

cles (AUVs): (1) data gathering from an underwater sensor network equipped

with acoustic communication and (2) autonomous inspection of the submerged

portion of a ship hull. For the first problem, I present path planning met

hods that extend algorithms for variants of the Traveling Salesperson Probl

em (TSP) and show how these algorithms can be integrated with realistic aco

ustic communication models. For the second problem, I discuss techniques f

or constructing watertight 3D meshes from sonar-derived point clouds and in

troduce uncertainty modeling through non-parametric Bayesian regression. Un

certainty modeling provides novel cost functions for planning the path of t

he robot that allow for formal analysis through connections to submodular o

ptimization and active learning. Such theoretical analysis provides insight
into the underlying structure of active sensing problems. Finally, I pres

ent experiments that demonstrate the high performance of the proposed solut

ions versus the state of the art in robot path planning.

Speaker Bio:

Geoffrey A. Hollinger is a Postdoctoral Research Associate in the Robotic
Embedded Systems Laboratory and Viterbi School of Engineering at the Unive

rsity of Southern California. He is currently interested in adaptive sensin

g and distributed coordination for robots operating with limited communicat

ion. He has also worked on multi-robot search at Carnegie Mellon University

, personal robotics at Intel Research Pittsburgh, active estimation at th

e University of Pennsylvania''s GRASP Laboratory, and miniature inspection
robots for the Space Shuttle at NASA''s Marshall Space Flight Center. He r

eceived his Ph.D. (2010) and M.S. (2007) in Robotics from Carnegie Mellon U

niversity and his B.S. in General Engineering along with his B.A. in Philos

ophy from Swarthmore College (2005).