8/3/2019 CEM Spring 2010

1/36

1 Spring 2009

0

MAGAZIN

LivingMachines:geneticaLLy

engineeringsoLutions to win

8/3/2019 CEM Spring 2010

2/36

2 Spring 200932 Spring 2010

The M.Eng. degree at Cornell is a unique

and wonderful opportunity because it can

be done anywhere around the world with

access to the nternet. This option has given

me exibility around my work schedule and

travel, yet does not sacrice the quality of

execution or caliber of education.

John Mar, Deputy Program Manager, BallisticMissile Defense, Lockheed Martin

www.systemseng.cornell.edu

Earn a CornEll M.Eng. in Demand for systems engineers is soaring. Whatwas once a niche job has become commonplacefrom medical device makers to automobilemanufacturers.

The Cornell Systems Engineering Master of

Engineering program reveals the systemsengineering process throughout an organizatioOur students learn to better designandmanagecomplex systems that evolvethroughout a life cycle. They become engineeriprofessionals capable of making trade-offdecisions in high-risk environments and creatinglobally optimized integrated solutions.

Cornells program is an important way for

our engineers to further develop their skills

and knowledge. The learning they bring back

ultimately improves the quality, creativity, and

competitiveness of Applied Materials products

and services.

Liang Y. Chen, Corporate VP and GM, Alternative EnerProducts Group, Applied Materials

online.

8/3/2019 CEM Spring 2010

3/36

Cnell engneeng 1

Living Machines: Geneticallyengineering solutions to win.By Dan Tuohy

8

12

Problem Solver: The pathless traveled leads to sciencesuperstardom for a oncestruggling ChemE.By Lauren Cahoon

Educating Innovators: KesslerFellows Program offers uniqueentrepreneurial instruction toengineering students.By Bridget Meeds

18

Features

DEPTMETS

22

contents

Killer Robots:The promise and perils of articial intelligence.By Michael Gillis

CELL EGIEEIG MGzIE/SPIG 2010

HMETHE

Pro. SheilaHemami

31Q& ITH

MKSPECE

162ES

28PEPLE

ProvIDED/STuroSnErPHoToG

raPHy

8/3/2019 CEM Spring 2010

4/36

2 ng 2010

NIH funds $13 million cancer center

The ational Cancer Institutehas funded the new Centeron the Microenvironmentand Metastasis, which willbe headquartered at Cornell.The center will focus on usingnanobiotechnology andother related physical scienceapproaches to advance researchon cancer.

ur center will be organiedto unravel cancers complexityusing methods derived from thephysical sciences and engineering to further understand howcancer travels through the humanbody, said Harold Craighead,Cornells C.. Lake Professor ofEngineering, director of Cornellsanobiotechnology Center,and the principal investigatorand director of the new center.The research may help identifynew drug possibilities to inhibit

metastasis and tumor growth.The new center is one of 12

new research centers across thenation announced ct. 26 by theinstitute. Cornells grant is for $13million over ve years.

Cornell will serve as the leadinstitution in a partnership witheill Cornell Medical College inew ork City and the niversityat Buffalo. Barbara Hempstead,professor of medicine andco-chief of the Division ofHematology and Medicalncology at eill Cornell MedicalCollege, will serve as the seniorco-investigator.

ationally, the 12 centers willbring a new cadre of theoreticalphysicists, mathematicians,chemists, and engineers to thestudy of cancer. During theinitiative, the Physical Sciencesncology Centers will take new,

nontraditional approaches tocancer research by studyingthe physical laws and principlesof cancer; evolution andevolutionary theory of cancer;information coding, decoding,transfer and translation incancer; and ways to de-convolutecancers complexity.

Cornells center will focus onthree key projects:

Examining physio-chemicaltransducers and their rolein tumor angiogenesis, ledby Claudia Fischbach-Teschl,Cornell assistant professor ofbiomedical engineering, andivek Mittal, associate professorof cardiothoracic surgery at eillCornell;

Physical and chemical cuesin tumor cell migration, led byCynthia einhart-King, Cornellassistant professor of biomedical

engineering, and ParaskeviGiannakakou, associate professorof pharmacology in medicine ateill Cornell; and

dhesion of tumor cells in thevascular microenvironment,led by Michael King, Cornellassociate professor of biomedicalengineering, and Davidanus, professor of medicineand urology and co-chief ofhematology and medicaloncology at eill Cornell.

ltimately, through coordinateddevelopment and testing ofnovel approaches to studyingcancer processes, the network ofcenters is expected to generatenew bodies of knowledge, andidentify and dene criticalaspects of physics, chemistry, andengineering that operate at alllevels in cancer processes.Blaine P. Friedlander Jr.

For years, researchers havestruggled to understand howtumors create the blood vesselsthat facilitate tumor growth, sothat one day, they can learn toblock the development of thesevessels and starve the tumor.

ow, with federal stimulusfunding from the mericanecovery and einvestment ct,

Team investigates how to starve tumors

Endothelial cellsa thin layer ofcells that lines the interior surfaceof blood vesselsseeded on thesurface of an engineered 3-D tumormodel (proliferating tumor cells inred, endothelial cells in green, cellnuclei in blue).

Cornell researchers will createtiny 3-D models of tumors tomimic conditions necessaryfor tumor angiogenesisthedevelopment of vascular systemsby tumors. sing such models,the researchers will specicallystudy characteristics of bloodvessels that feed tumors andtransport stem cells, which, inturn, play a role in developing a

tumors vascular system.The research teamincluding

Claudia Fischbach-Teschl,assistant professor of biomedicalengineering; be Stroock andJeff arner, assistant professorsof chemical and biomolecularengineering; and ivekMittal, associate professor ofcardiothoracic surgery at eillCornell Medical Collegewillreceive $633,000 over two years.

The funding will supportthree graduate students and a

postdoctoral researcher to workon the project.

mong other things, theresearchers hope that developinga microuidic 3-D tumor modelwill allow them to examinehow such essential solubleelements as dissolved oxygen aretransported in blood to tumors.The researchers will also oodthe system with bone marrow-

derived stem cells that play a rolein tumor vasculariation.

e will look at how and wherethey incorporate and be able tostudy these cells qualitatively andquantitatively, said Fischbach-Teschl.

The researchers will also usemathematical modeling to studyunderlying molecular signalingpathways involved in tumorangiogenesis.Krishna Ramanujan

ProvIDED/SCoTTS.v

ErBrIDGE

8/3/2019 CEM Spring 2010

5/36

Cnell engneeng 3

ISS 1081-3977

olume 16, umber 1

Spring 2010

Cornell Engineering

Magazine is published

by the Cornell niversity

College of Engineering

Interim Dean

Christopher K. ber

Francis Bard Professor of

Materials Science and

Engineering

Associate Deanfor Administration

Cathy Dove

Executive Editor

Barbara L. Cain

Director

Engineering Communications

and Media elations

Editor

obert B. Emro

ssistant Director

Engineering Communications

and Media elations

Art Director

Clive Howard

ffice of Publications and

Marketing

Graphic Designer

Lorraine Heasley

ffice of Publications and

Marketing

Printer

Midstate Litho

Endicott, ..

Photography

ll photos by niversity

Photography unless

otherwise indicated

Editorial and

Business Offices

B1D Carpenter Hall

Ithaca, 14853-2201phone 607 255-3981

fax 607 255-9606

e-mail cornell_engr_mag@

cornell.edu

isit Cornell EngineeringMagazine online at

www.engineering.cornell.edu/magaine

2010 Cornell EngineeringMagazine

Printed on recycled paper.

03/10 ML 42M 100196

Beetle-inspired device could let humans walk on walls

Could humans one day walk onwalls, like Spider-Man? palm-sied device invented at Cornellthat uses water-surface tensionas an adhesive bond just mightmake it possible.

The rapid adhesion mechanismcould lead to such applicationsas shoes or gloves that stickand unstick to walls, or Post-it-like notes that can bear loads,according to Paul Steen, professorof chemical and biomolecularengineering, who invented thedevice with Michael ogel, aformer postdoctoral associate.

The device is the result ofinspiration drawn from a beetlenative to Florida, that can adhereto a leaf with a force 100 timesits own weight, yet also instantlyunstick itself. esearch behind thedevice was published online Feb.1 in Proceedings of the NationalAcademy of Sciences.

The device consists of a atplate patterned with holes, eachon the order of microns (one-millionth of a meter). bottomplate holds a liquid reservoir, andin the middle is another porouslayer. n electric eld appliedby a common 9-volt batterypumps water through the deviceand causes droplets to squeeethrough the top layer. The surfacetension of the exposed dropletsmakes the device grip anothersurfacemuch the way two wetglass slides stick together.

In our everyday experience,

these forces are relatively weak,Steen said. But if you make a lotof them and can control them,like the beetle does, you can getstrong adhesion forces.

For example, one of the

researchers prototypes was madewith about 1,000 300-micron-sied holes, and it can holdabout 30 gramsmore than 70paper clips. They found that asthey scaled down the holes andpacked more of them onto thedevice, the adhesion got stronger.They estimate, then, that a one-square-inch device with millionsof 1-micron-sied holes could holdmore than 15 pounds.

To turn the adhesion off, theelectric eld is simply reversed,and the water is pulled backthrough the pores, breaking thetiny bridges created betweenthe device and the other surfaceby the individual droplets.

The research builds onpreviously published work thatdemonstrated the efcacy ofwhats called electro-osmoticpumping between surfacetension-held interfaces, rstby using just two larger waterdroplets.

ne of the biggest challengesin making these devices work,Steen said, was keeping thedroplets from coalescing, aswater droplets tend to do whenthey get close together. To solvethis, they designed their pump toresist water ow while its turnedoff.

Steen envisions futureprototypes on a grander scale,once the pump mechanismis perfected, and the adhesivebond can be made even stronger.

He also imagines covering thedroplets with thin membranesthin enough to be controlledby the pump but thick enoughto eliminate wetting. Theencapsulated liquid could exert

Paul Steen and Michael Vogelssurface tension-based adhesivedevice with a Lego man payload.A video of a larger version of thedevice is at www.engineering.cornell.edu/cem.

simultaneous forces, like tinypunches.

ou can think about making acredit card-sied device that youcan put in a rock ssure or a door,and break it open with very littlevoltage, Steen said. Its a funthing to think about.

The research was fundedprimarily by the Defensedvanced esearch Projectsgency, and also by the ationalScience Foundation.Anne Ju

ProvIDED/MICHaELvoGEL

8/3/2019 CEM Spring 2010

6/36

4 ng 2010

ew research by

Thorsten Joachimsand obertKleinberg, associateand assistantprofessors of

computer science, respectively,aims to create search-enginesoftware that can learn fromusers by noticing which linksthey click on in a list of searchresponses, and how theyreformulate their queries whenthe rst results dont pay off.

The work is funded by a four-year, $1 million grant from theational Science Foundationunder federal stimulus funding,formally known as the mericanecovery and einvestmentct. The research will lead tomethods that improve searchquality without human guidance,especially on specialied ebsites such as scientic or legalcollections or corporate intranets.

Joachims believes the work willhave long-term benets for theeconomy, invigorating the marketfor high-quality and focused

search software. I think thereis a potential for commercial

impact, improving quality and

productivity, he said. In the shortterm, the project will fund at leasttwo Ph.D. students for four years,and provide research positions forundergraduate students.

s a demonstration, theresearchers plan to create a newsearch engine for the physicsarXiv eb site at Cornell, whichcontains thousands of papersin physics, mathematics, andcomputer science, and possiblyfor other specialied collections.

In several ways, providingsearch for small collectionsis more difcult than for thewhole Internet. Google, ahoo!,and Microsoft can spend a lotof manpower on engineering agood ranking function for theInternet. For small collections,this has to happen automaticallyvia machine learning to beeconomical, Joachims explained.

Search is not a one-sie-ts-all business: People searchingspecialied collections mightuse the same words in verydifferent ways. Is uncertainty,

for example, about thelocation of subatomic particles,

Stimulus funding to help search engines learn on the job

ProvIDED

There is a trade-off. n theone hand, you want to presentthe best ranking you know sofar, Joachims explained. n theother hand, the search engine hasto do a bit of experimentationto be able to learn even betterrankings in the long run. Thekey is to balance the tradeoffbetween presentation andexperimentation in an optimalway.

This trade-off is similar to whata gambler faces in a casino andis called a multi-armed banditproblem. hen playing a row ofslot machines, each play givesyou new information about howmuch that machine pays, butalso costs you a quarter. The trickis to eliminate some machineswhen youre sure they wontpay off without spending morethan necessary. Kleinbergs workon algorithms for solving suchtrade-off problems will be key

to making search engines learneffectively.Bill Steele

career choices, investment

opportunities, or romance?The key idea is have a searchengine that gets better just bypeople using it, Joachims said.He and his collaborators havealready created a search enginecalled smotthe name is a playon learning by osmosisthatdraws on extensive research bycomputer scientists in machinelearning. The problem the newresearch will address is thatwhat the machine learns maybe biased by the way it displaysresults.

Eye-tracking studies done incooperation with Geri Gay, theKenneth J. Bissett Professor andchair of Cornells Departmentof Communication, have shownthat absence of a click on a resultat, say, the 11th position on thelist of returns may mean thatthe result did not t the usersinformation need, but it may alsomean that the user had given upscanning the list that far down.To get reliable feedback fromclicks, the search engine needs to

shufe the order in which resultsare returned.

The Cornell Society of omenEngineers received the Gold Levelward for utstanding CollegiateSection at the national SEconference held ct. 1517 in LongBeach, Calif.

The top national honor forSE collegiate chapters, theaward recognied Cornell SEsoutstanding achievements in awide range of areas, includingplanning and administration,communication and memberrecruitment, regional/nationalparticipation, education andoutreach, inclusiveness anddiversity, and alignment with thenational SE mission.

Cornell SE is led by co-presidents Kristie esetco 10 E

Society of Women Engineers takes home national award

Thorsten Joachims

and Emily Swarr 10 ChE.The colleges freshman class

is 37 percent female, an all-timehigh.

Anne Ju

SWE Co-presidents Emily Swarr 10ChE (left) and Kristie Resetco 10 ORE(right) holding the award at theSWE National Conference.

ProvIDED/JEnnIFErDouGHTy

ProvIDED/JEnnIFErDou

GHTy

Alexandra Woldman 10 CE helps children make a putty-like elastic polymeras part of a SWE event at the Ithaca Sciencecenter in November.

8/3/2019 CEM Spring 2010

7/36

Cnell engneeng 5

In early ovember, Cornell wasnamed one of 10 seminalists

in the rst Multi-utonomousGround-robotic InternationalChallenge (MGIC), co-sponsoredby the .S. Department ofDefense and the ustralianDefense Science and Technologyrganiation. The competitionrequires teams to build variousrobots that can accomplish aseries of tasks like scanning fordangerous objects, protectingcivilians, and mapping forobstructionsall with minimalhuman supervision.

s a seminalist, Cornellreceived $50,000 to continuedeveloping its robots andpreparing for another roundof cuts in June. Five nalistswill compete in ustralia inovember 2010.

ssistant Professor HadasKress-Gait and ssociateProfessor Mark Campbell, bothin mechanical and aerospaceengineering, co-lead the project

MAGIC robots make cut Down Under

annEJu/CornELLCHronICLE

Members of Cornells Multi-Autonomous Ground-robotic InternationalChallenge team in the Autonomous Systems Lab with two of their robotprototypes. From left: graduate student Danelle Shah; Daniel Lee 10 ChE; labmanager Aaron Nathan 06, M.Eng. 09 CS; Chuck Yang M.Eng. 10 ECE; andgraduate student Mark McClelland.

and moving objects of interest.They will also be required to have

mapping capabilities to create arecord of the area.

MGIC aims to nd ways inwhich one to two operators cancontrol 50 robots to magnify

their effectiveness, said graduatestudent Mark McClelland. The

swarm of robots can do theirthing, while the humans do thehigh-level stuff that humans aregood at.Anne Ju

as part of the utonomousSystems Lab, which produced

Cornells DP Challengeautonomous vehicles.

e took an interest inthe researchnot just thecompetition, Kress-Gait saidof MGIC. f course we wantto win, but thats not the maindrive.

The Cornell MGIC team,consisting of about 20 peopleincluding faculty, graduatestudents, staff, and someundergraduates, has four robotprototypes with more possible

in the future. They are built onSegway platforms, and each willeventually be equipped with GPSunits, laser range nders, cameras,and inertial measurement units,which detect acceleration.

For the competition, therobots will operate in an urbanenvironment, both indoors andoutdoors. In each environment,the robots must nd, identify, andcorrectly respond to stationary

CEAA creates endowment

The CornellEngineeringlumni

ssociation recently created theCE Legacy Endowment Fundwith a donation to the collegeof more than $180,000. Incomefrom the fund will be used forawards recogniing outstandingachievement by faculty and

students. It also will sponsorgrants to student teams.

e have long rewarded facultyand students for outstandingachievement, said CEPresident Bill Bruno 69 CE,M.Eng. 71. The fund will allow usto further that goal.

Bruno said a grant the CEprovided to the Cornell SolarDecathlon team, which competedlast ctober in ashington,D.C., was an example of the kindof team sponsorship the fund

would support. Several additionalgrants are planned for this year,he said.

The donated money camefrom fees collected for lifetimemembership in the association,formerly known as the CornellSociety of Engineers. In 2007,the CE board of directorsvoted unanimously to eliminate

membership dues, makingall alumni members free ofcharge. The CE wants toincrease alumni involvement,explained Bruno. Participatingin CE activities will enhanceyour connection to Cornell, itsstudents and faculty, and otheralumni.

CE-sponsored conferencesare one way to get involved,said Bruno. In ovember, theassociation helped organie aconference in Philadelphia titled

t the Leading Edge: ImprovingLives Through BiomedicalEngineering. Speakers includedClaudia Fischbach-Teschl,assistant professor of biomedicalengineering, and C.C. Chu, theebecca Q. Morgan 60 Professorof Fiber Science and pparelDesign in the College of HumanEcology. They were joined by

Cornell alumnus and trusteeDavid . Fischell 75 EP, M.S. 78P, Ph.D. 80, founder of ninebiomedical device companies.

ll who attended had agreat time exploring newthoughts, meeting old friends,and making new friends,said Bruno. Participants saidthat reconnecting with otherengineering alumni was aswonderful as the program itself.

CE and the CornellEntrepreneur etwork are co-

sponsoring the next conference,Sustainable Energy: Investing inur Future, which will take placein Boston on June 17.

Besides giving alumni away to get together, the CEprovides many ways to connectwith Cornell engineeringstudents, Bruno said, such asspeaking on campus, mentoring

students, assisting job seekers,and sponsoring Master ofEngineering projects. e alsohost dinners for Cornell co-opstudents in locations away fromIthaca so that they may meetalumni and their fellow students,he said.

ews updates, details of futureconferences, and informationon how to become involved inthe CE can be found at www.engineering.cornell.edu/ceaa.Robert Emro

8/3/2019 CEM Spring 2010

8/36

6 ng 2010

Lara Estroff, assistant professorof materials science and

engineering, and colleagues havetaken a detailed look at the waymanmade calcite crystals, similarto those found in limestone, growin tandem with proteins andother large molecules, as they doin biological structures like shells.

They reported their ndings inthe ov. 27 issue of the journal

Science.e knew the organics were

in there, but what no one hadbeen able to do up until nowwas actually see what thatorganic-inorganic interfacelooked like, said Estroff, whoselab focuses on the synthesis andcharacteriation of bio-inspiredmaterials.

Estroff and graduate studentHanying Li grew samples ofcalcite in a hydrogel calledagarose that mimics the way

calcite grows in living things.In previously published work, Liand Estroff had determined thatthis gel environment made thecrystals grow very differentlythan in solution.

Research offers clues to how shells grow

ESTroFF/MuLLErLaBS

This image shows the 3-D nanoscale structure of the inside of a syntheticcalcite crystal, which was grown in an agarose gel. In this environment, thecrystal grows around the polymeric bers. The structure gives insight intothe formation of single crystals by biological organisms.

ith a bit of leverage, Cornellresearchers have used a verytiny beam of light with as little

as one milliwatt of power tomove a silicon structure up to 12nanometers. Thats enough tocompletely switch the opticalproperties of the structure fromopaque to transparent, theyreported.

The technology could haveapplications in the designof micro-electromechanicalsystems (MEMS)nanoscaledevices with moving partsandmicro-optomechanical systems(MMS) that combine movingparts with photonic circuits, saidMichal Lipson, associate professorof electrical and computerengineering.

The research by postdoctoralresearcher Gustavo ieder-hecker, Long Chen M.S. 08 ECE,Ph.D. 09, lexander Gondarenko,M.S. 09 P, Ph.D. 10, and Lipsonappeared in the journal Nature.

Light can be thought of as astream of particles that can exerta force on whatever they strike.The sun doesnt knock you offyour feet because the force is

Scanning electron micrographof two thin, at rings of silicon

nitride, each 190 nanometers thickand mounted a millionth of ameter apart. Light is fed into thering resonators from the straightwaveguide at the right.

Small optical force can budge nanoscale objects

In collaboration with associateprofessor of applied andengineering physics David Mullerand physics graduate studentHuolin Xin, the researchersprepared their crystals withFocused Ion Beam technology,which uses high-energy ions toslice samples thin enough for an

electron beam to pass throughfor imaging.

Scanning transmission electronmicroscopy methods developedby Xin and Muller revealeddetailed, three-dimensional

pictures of the internal structureof calcite crystals grown in thegel. They found that the crystalstrap large molecules by growingaround them.

Studying this complex naturalprocess may be a key step towardgiving materials scientists likeEstroff clues on how to make

and manipulate nature-inspiredcomposite materials. pplicationscould range from electronics tophotovoltaics to completely newclasses of materials.Anne Ju

very small, but at the nanoscale itcan be signicant. The challengeis that large optical forces are

required to change the geometryof photonic structures, Lipsonexplained.

But the researchers were ableto reduce the force required bycreating two ring resonatorscircular waveguides whosecircumference is matched to amultiple of the wavelength of thelight usedand exploiting thecoupling between beams of lighttraveling through the two rings.

beam of light consists ofoscillating electric and magneticelds, and these elds can pullnearby objects toward the centerof the beam, a microscopicequivalent of the way staticelectricity on clothes attracts lint.This phenomenon is exploitedin optical tweeers used byphysicists to trap tiny objects.

hen light travels through awaveguide whose cross-sectionis smaller than its wavelength,some of the light spills overand, with it, the attractiveforce. So parallel waveguidesclose together, each carrying

a light beam, are drawn evencloser, rather like two streams ofrainwater on a windowpane thattouch and are pulled together bysurface tension.

The researchers created astructure consisting of twothin, at silicon nitride ringsabout 30 microns in diameter

mounted one above the otherand connected to a pedestal bythin spokes. The ring waveguides

are three microns wide and 190nanometers thick, and the ringsare spaced one micron apart.

hen light at a resonantfrequency of the rings, in thiscase infrared light at 1533.5 nm,is fed into the rings, the forcebetween the rings is enough todeform the rings by up to 12 nm,which the researchers showedwas enough to change otherresonances and switch otherlight beams traveling through therings on and off. hen light inboth rings is in phasethe peaksand valleys of the wave matchthe two rings are pulled together.hen it is out of phase, they arerepelled. The latter phenomenonmight be useful in MEMS, wherean ongoing problem is that siliconparts tend to stick together,Lipson said.

n application in photoniccircuits might be to createa tunable lter to pass oneparticular optical wavelength,iederhecker suggested.Bill Steele

Michal Lipson

Lara Estroff

CornELLnanoPHoTonICSGrouP

8/3/2019 CEM Spring 2010

9/36

Cnell engneeng 7

t the end of ctober, theCornell 100+ MPG Team learnedit had made the short listof teams competing in nextyears $10 million ProgressiveInsurance utomotive X Prie,an international competition inwhich cars that get 100 milesto the gallon or more will gohead-to-head in a series of races.Cornell is the only university

team represented in themainstream vehicles class. Thenext phase of the competition

100+ mpg car hits the road for a test-drive CEM goespaperless

Alumni andriends o thecollege can now sign upto receive an electronicversion oCornellEngineering Magazineinstead o a printedcopy.

The college isproviding thispaperless option toreduce its impact onthe environment andsave money spent onprinting and mailingcosts.

Anyone who wouldlike to receive an e-mailnotifcation when wepublish a new issueoCornell Engineering

Magazine onlinecan do so at www.

engineering.cornell.edu/paperless.

is the technical qualiers, whichbegin this spring.

hen the team got the goodnews, the car was starting tocome togetherits red paintwas emblaoned with a list ofsponsors on the side; the gutsof the car were talking to eachother; the electrical drive trainwas spinning the wheels, and thebraking system was functioning

fully.The 100-mpg goal will be metwhen the cars generator, the

backup power for the lithium-iron batteries, is installed.

British company, Evo Electric, hasdonated a $20,000 generatorthats being programmed forthe car, replacing a bulkier, lesspowerful one the team had beenusing. The new, pancake-stylegenerator is only about 6 incheslong as opposed to the old 20-inch generator.

The team has been helpedalong the way by other sponsors,including the lithium-ion batterypack from Changs scendingCo., a Taiwanese company, and a

$20,000 cash donation from itsparent company, Formosa Inc.

The team conducted their rstdrive test behind hodes Hallbefore moving to B Lot for morereal-world testing conditions.

ntil next May, when the carswill be complete and the racesbegin, the team will need tosubmit to competition judgesseveral detailed technical reportson the car, including informationon the fuel system, the safety ofthe battery packs and the crash-

worthiness of the vehicle.Anne Ju

How far does thetrain go?Professor Andy Ruina, Mechanical and

Aerospace Engineering

A train is pulled by a cable that wrapsaround a cylinder that is parallel to the

tracks. How ar can the cylinder pullthe train?

Solve this to win Cornell Engineering gear

Te cable can pull no more once it is orthogonal to the top line of the

cylinder. Te top of the cylinder is level with the cable attachment

point on the train and a distance d from it. Te cable is inextensible,

always taut and does not slip once it is on the cylinder. Te cylinder

has radius R. (a) At the start of the process the cable AB makes a 30

degree angle with the top of the cylinder. (b) Limiting case: at the start

of the process A is innitely forward of the train. Hint: the answers

have a simple form. Question unclear? Ask Andy: ruina@cornell.edu.

t

Enter your answers to both (a) and (b) by ug. 1at www.engineering.cornell.edu/brainteaser. ewill draw 12 winners from the correct entries. Theywill receive a Champion sweatshirt (ell contact

you for siing.), a Garland gift pen, a stainless steelwater bottle, or one of nine Cornell Engineering cardecals. e will announce the winners in the Fall

2010 issue.The full line of Cornell Engineering gearcan be found at www.cornell.engineering.edu/gear.

ProvIDED

cylinder

no slip

R

B d

cable

xedtracks

ProvIDED

motor

A

8/3/2019 CEM Spring 2010

10/36

8 ng 2010

Problem

SolverBy Lauren CahoonThe pth e tveed ed to cece upetdomfo oce tu ChemE.

8/3/2019 CEM Spring 2010

11/36

ithin the scientic community, and even

outside of it, obert Langer 70 ChE is something

of a celebrity. Colleagues toss around the

word genius when referring to him. Entire

elds of research are considered his brainchildren. Life-saving

technologies are on the market due to his ingenuity. et, if you

time-traveled back to 1967 to get a glimpse of Langer as a Cornell

undergrad and expected to see a science prodigy acing his

prelims, youd be surprised. My rst three terms were pretty bad,

says Langer. I dont think my average was above a B minus.

This disarming revelation may bring hope to those Cornell

engineering students who nd their coursework less than

breey. Despite the rocky start, Langer went on to get his Sc.D.

at the Massachusetts Institute of Technology where he wouldlater become a professor in bioengineering, running the largest

academic biotechnology lab in the world. He would also garner

more than 170 major awards, including the 2008 Millennium

Prie, (the worlds largest technology prie), the 2006 nited

States ational Medal of Science, and the Charles Stark Draper

Prie (the equivalent of the obel Prie for engineers). nd he

has returned to Cornell many times as a guest lecturermost

recently, he spoke to engineering students last fall about

biomaterials and biotechnology, focusing on the two elds he

revolutionied: tissue engineering and drug delivery systems.

How did he turn a dismal rst three semesters into a science

superstar career? fter a year and a half of falling asleep in large

lecture classes and not getting anything out of it, I thought, hy

am I even bothering to go? Langer says. He realied he needed

to change his hands-off approach and began attending ofce

hours and pushing himself in labsand focused on solving lots

of problem sets. For Langer, this was the key to success. It was a

self-observation after three terms of doing crummy, I realied I

was good at solving problems, he says.

His problem-solving skills soon shot him to straight- status,

and he thrived as a T in chemical engineering classes. hat lay

in store after graduation, however, was a big blank, says Langer.

Even his choice of majors was a whimsical decision. In my rstyear the only thing I did well in was chemistry, and I was in the

engineering school, so I picked it. I dont think I even knew what a

chemical engineer actually did.

s it turned out, most chemical engineers in his class were

taking jobs in the oil industryan idea that held little appeal

for Langer. I always had the idea that I wanted to do something

good for the world, he says. It was this noble, yet vague desire

that led Langer to the unusual decision that would jumpstart his

careerapplying for and winning a post-doctoral position in the

lab of Harvard Medical School vascular biology researcher Judah

MIT Professor obert Langer 70 ChE and fellow faculty memberJeffrey Karp display an adhesive they developed that was inspired bythe gecko and may have medical and surgical applications.

Cnell engneeng 9

ProvIDED/DonnaCovEnEy/MIT

8/3/2019 CEM Spring 2010

12/36

10 ng 2010

Langer described his work to Cornell faculty and students at a seminarlast September.

Folkman. It was an unusual move for an engineer, Langer admits,

but a fortuitous one. There, Langer and Folkman tackled the issue

of drug delivery.

Folkman wanted to develop a type of drug known as an

angiogenesis inhibitor, which blocks the growth of blood vessels

that nourish tumors. However, these drugs cannot be taken orally

and are ineffective when injected. They needed a new way to get

these drugs to tumorsand Langer believed this could be done

using porous polymer microspheres that would slowly releasethe drugs directly at the tumor site after being injected. et the

idea of a porous polymer that would allow the medications

large molecules to pass through it was deemed impossible

by the scientic community. This didnt fae Langer, who was

determined to nd a solution, despite the odds. Judah Folkman

and I found about 200 ways to get it to T work, Langer

says with a smile. et persistence paid off. ow drug delivery

systems, based in large part on his work, are used to treat cancer,

schiophrenia, and alcoholism. Clinical trials are under way to use

such systems for diabetes and substance abuse.

Since then, Langer has continued to forge meaningful medical

advances, including tissue engineering, a eld that he actively

created by discovering the use of polymer-based scaffolds on

which human cells could be placed and encouraged to replicate

and ourish. Langers research in this eld has since developed

exponentially, with scientists now having successfully grown

human skin, corneas, bladders, cartilage, bone, and blood vessels,

while spinal cord, liver, pancreas, intestine, and ureters have been

successfully grown in animal studies.

For all the accomplishments, Langer is remarkably down to

earth. Speaking with him face to face during his visit to Cornell

last September, the scientist was unassuming, answering

interview questions as if for the rst time, rather than theve-hundredth. Hes happy to recall memories from his Cornell

undergrad days, from sledding down Libe Slope on cafeteria trays

(Do they still do that anymore?) to a now long-gone dining

establishment Langer and his friends would call The Barf Bar.

Some people are generally successful in their career, but

theyre not very easy to get along with, but Bobs not like that,

says Mike Shuler, the James M. and Marsha McCormick Chair of

the Department of Biomedical Engineering, whos known Langer

for decades. lex Klibanov, MIT ovartis Professor of Chemistry

and Bioengineering, who met Langer 30 years ago when they

both joined the faculty at MIT, agrees. e have been good friends

for years, so Im not sure Im objective, says Klibanov, but there is

a wide consensus in the eld that hes a very nice person.

Both Shuler and Klibanov claim this has worked greatly to

Langers advantageit fosters a highly successful lab of graduate

students and post-docs that create rst-rate research. ith

close to 1,100 scientic papers and 750 issued and pending

patents to their credit, Langer lab members have a laundry list of

discoveries that aim to change the face of medicine. Some of the

more recent ones include work on creating long-term delivery

systems for insulin, growth factors, gene therapy agents, vaccines,

and anti-cancer drugs. ther work focuses on drug delivery

systems that can move proteins or genes across anatomical

barriers such as the intestine, lungs, blood-brain barrier, and the

skin. nd, as mentioned before, tissue engineering is another keyfocus in the labwith the famed ex-songbird Julie ndrews as a

potential beneciary of their efforts. By injecting a synthetic, gel-

like material into scarred or damaged vocal cords, Langer and his

team believe they may someday be able to restore the voices of

singers like ndrews.

hile Langer has done a lot for the medical eld, his work

hasnt stopped there. He recently teamed up with a hair product

company known as Living Proof to tackle the omnipresent

problem of friy hair. The result was o Fri, a product

composed of a compound known as polyourester that coats hair

follicles and prevents moisture from getting at the hair strand

and creating a fri response. Beauty magaines and fri-haired

customers have raved about the product since its release earlier

this year.

ne particularly faithful customer happens to be Langers

daughter, Susan 13, who got to work at Living Proof during

its early stages doing both science and marketing. Susan, a

chemistry major at Cornell who hopes to work in science and

business one day, seems to be a chip off the Langer block.

et she says her father never pushed her into scienceand

that dinnertime conversations at the Langer household dont

8/3/2019 CEM Spring 2010

13/36

The C55 ure Dynamicsgenerator charges the batteriesfrom the diesel engine.

Cnell engneeng 11

necessarily focus on her fathers high-prole science work but on

every-day stuff, says Susan. e talk about everybodys lives at

the dinner table. She also mentions fond memories of attending

weekly summertime Langer lab softball gamesa well-loved

tradition among Langer group members.

Hes very, very good at working with a variety of people, adds

Shuler, he elicits a lot of loyalty, and people feel theyre doing

something important. This may in part be due to Langers

unfailing policy to make time for whoever requests to see or talk

to himsomewhat unusual for a scientist of Langers stature.

However, thats just showing people respect, says Langer.

If someone wants to talk to me, I will make time for them.

That sense of respect seems to have an empowering effect.

f the roughly 500 students and post docs to come through

the Langer lab, 200 of them are executives at pharmaceutical

and biotech companies, and 200 of them are now professors,

including David Putnam, an associate professor of chemical and

biomolecular engineering at Cornell. Putnam was a postdoc in

the Langer lab from 1996 to 2000 and vows the experience was

instrumental in his success. I was in an environment where you

could dream bigBob didnt micro-manage, says Putnam. It

was an academic candy store. Langer says it is these successful

careers, like Putnamsand not the countless pries, papers, and

patentsthat he is most proud of.

s for those who aspire to follow Langers footsteps, consider

not going down a conventional path, says Langer. hatever

else you do, think about what impact is the greatestis it the

road everyone else is taking? nd, after years of going down the

path less traveled, Langer now knows better than to listen to

naysayers. lot of people may tell you its impossible, he says.

But theres not that much thats impossible. Its worth trying it if

you believe.

Whatever else you do, think about What impact is greatestis it the road

everyone else is taking? a lot of people may tell you its impossible. but

theres not that much thats impossible. its Worth trying it if you believe.

robert langer

8/3/2019 CEM Spring 2010

14/36

12 ng 2010

livingmachineSgEnETiCally EnginEEring slTins T win

Ben Cammarata 11 BE and Malinka alaliyadde 12 MSE creating BioBricks for the competition.

8/3/2019 CEM Spring 2010

15/36

Cnell engneeng 13

screen ashed an image of an impoverished man, hunched over and suffering from

prolonged exposure to toxic heavy metals. Malinka alaliyadde 12 MSE stepped up to

the front of the hall at the Massachusetts Institute of Technology in Cambridge, Mass., to

identify the culprit: cadmium poisoning.

Ingestion of cadmium-contaminated water, whether from industrial and mining

operations, sewage, or poor irrigation practices, can cause any number of life-threatening

conditions and lead to bone fractures and severe brain and kidney damage.e know cadmium is bad for us, alaliyadde says. In developing countries it remains

an issue. ne of the biggest problems in developing countries is the consumption of

cadmium-contaminated crops.

He and ve others on one of Cornells newest student teams have created an answer: a

biosensor to detect cadmium. The team of undergraduates put their research to the test

at MIT in the 2009 International Genetically Engineered Machine competition, better

known as iGEM, and earned a brone medal.

In presenting their work, the teammates switched up, each describing part of the

problem and their solution: attaching proteins that glow when activated to the cadmium

response promoters found in the soil bacteria Bacillus subtilis to synthesie a whole cell

cadmium biosensor. bacterial biosensor of cadmium can be much less expensive and therefore more

readily available in third-world countries, says Xing Xiong 10 Chemistry, Cornells team

leader.

The opening presentation at MIT took approximately 30 minutes, but the team began

training for the competition a year before. Xiong, alaliyadde, Bernard Cammarata 11 BE,

lyssa Henning 11 BE, Matthew Hall 10 BE, and Kevin Cheng 10 Bio worked with advisers

Carl Batt, Maki Inada, John D. Helmann, and Xiling Shen.

The expertise is indicative of the multidisciplinary approach. Batt is a professor of

food science. Inada is an adjunct assistant professor of molecular biology and genetics.

Helmann is a professor of microbiology. Shen is an assistant professor of electrical and

computer engineering.

The premiere undergraduate synthetic biology competition, iGEM requires student

teams to begin working with a standard kit of biological parts, known as BioBricks,

at the beginning of summer. They use the parts and parts of their own design to build

biological systems and operate them in living cells.

It began in 2003, designed as an effective and motivational teaching method, with a

month-long independent studies course at MIT. Students designed biological systems to

make cells blink. The rst competition, in 2004, was among ve teams. It grew to 13 teams

in 2005, 32 teams in 2006, 54 teams in 2007, and 84 teams in 2008.

Projects have ranged from banana and wintergreen-smelling bacteria to arsenic

biosensors and buoyant

bacteria.

The iGEM philosophy is getand give, said iGEM director

andy ettberg as he welcomed

the record number of 2009

participants to MIT. They

numbered 1,700 from 112 teams

and came from 26 countries in

pursuit of research glory and a

chance to change the world.

iGEM points to the impressive

By Dan Tuohy

ECE ssistant Professor Xiling Shen advises the team.

ProvIDED/aL

ySSaHEnnInG

8/3/2019 CEM Spring 2010

16/36

14 ng 2010

student team accomplishments, often reached in a short span of

time, which may lead to advances in medicine, energy, and the

environment.

Sponsors in 2009 included Life Technologies, a global

biotechnology tools company; GeneT, a world provider

of synthetic genes; the ational Science Foundation; The

Mathworks; and the Federal Bureau of Investigation, through its

Biological Sciences utreach Program.

Its just been growing like mad, says Meagan Liarao,

assistant director of iGEM. Its a new eld, so that kind of attracts

people.

Students say it is also fun, all this hands-on work and research

and the professional interaction.

Theyre the ones driving the whole project, Liarao says. The

students feel like they own the project.

The Cornell team was founded by aweed Paya 09 BE ECE,

M.Eng. His interest led to exploration in 2008. Students conferred

and brainstormed, discussing with faculty how best to launch a

team.

I was already interested in this whole synthetic biology eld,

along with electrical engineering that I was doing, Paya says.

fter reading up on iGEM, he says he was motivated by the

MineSweeper and autonomous underwater vehicle teams at

Cornell, and the vacuum in the synthetic biology eld.

There was a lot of interest when we sent out the initial calls,

says Paya. There was a lot of interest and there was really

nothing available. It was the perfect opportunity.

The biggest challenge? Finding lab space on campus, the

students say during interviews after their initial presentation at

MIT.

The multidisciplinary challenges remain an attractive

component, says Henning. Malinka and I didnt have experience

in the genetic engineering, she says. So there was a bit of a

learning curve there.

Hall says the iGEM philosophy, that the competition advances

each year based on all the teams work, is compelling.

I feel like now that Ive done the projects, I can understand the

lectures a lot better, he says. Last time we split up, saw probably

half of the lectures, and took all these notes. e really gured

out a couple weeks later that we had no idea what had just

happened.

fter their performance at MIT, Xiong says the questions from

I know a lot about the techniques, how to do the cloning and things like that, but Idont think about circuits. ... So having the engineers with me, from that perspective,

really enriches the whole project. Maki Inada

lyssa Henning 11 BE culturing B. subtilis in tubes. The bacteria grow overnight and can be used in experiments the next day.

ProvIDED/aLySSaHEnnInG

8/3/2019 CEM Spring 2010

17/36

Cnell engneeng 15

ProvIDEDjudges were reasonable, centered as they were on design and

purpose. e wanted to do something that we could accomplish

in a short period of time, he says. e knew that it was our

rst year so we could have some trouble, so we tried to think

of something simple and we had seen some people do some

biosensors before.

Those biosensors were different, of course. Henning says

the teams early brainstorming produced biomedical and

environmental ideas. nother reason why we chose to go with acadmium biosensor is that Bacillus subtilis, the bacteria we work

with, naturally took in metals that weve got, she says. ll the

other teams use E. coli, the workhorse of bacteria. But E. coli does

not naturally take in metals, so it would have to be manipulated

more to produce the desired biosensor.

hile biosensors are pretty well established, adviser Shen noted

the Cornell concept is a cheap way to create them, something of

paramount importance in developing nations. These provide a

very, very convenient way, he says.

Shen says the iGEM competition has several educational

benets, starting with its interdisciplinary scope. Because it

requires different majors with complementary skills to work

together, Shen and the other team advisers have created a special

course, cross-listed in multiple departments, to recruit and

prepare students for future iGEM competitions.

Shen says organiers try to make the competition as realistic as

possible. In some course work, there can be a disconnect between

studies and occupational reality, so it can be a unique experience

for undergraduates.

lso, he says, the organiers try, in a way, to show that

research is cool.

Cheng called Cornell progressive for its support of the new

team. Its really very forward-looking as education, especiallybiology, he says. In 50 years, itll be unrecogniable with what

we see today.

Teams continue to draw talent from across the spectrum.

Liarao pointed out that several teams had social science majors,

including a couple of artists, among their members. ne of those

teams was Cambridge niversity, the grand winner in 2009.

Inada, the molecular biologist, spoke of the valuable

collaboration in confronting a growing eld and emerging

technologies.

I know a lot about the techniques, how to do the cloning

and things like that, but I dont think about circuits, Inada says.

I dont think about how all the pathways connect together.

So having the engineers with me, from that perspective, really

enriches the whole project.

Interdisciplinary assets are visible on campus, as Inada

explained. The engineering and biology departments are not

really close to one another. But the new eill Institute for Cell and

Molecular Biology, a state-of-the-art center in newly constructed

eill Hall, is a kind of connective tissue.

They have bio engineers in that building that are sort of

helping to bridge and bring together the different disciplines,

she says. Its a really exciting time for that.

Members of the CGEM team stand in front of their poster at the 2009international iGEM jamboree. From left: lyssa Henning 11 BE, Matt Hall10 BE, Xing Xiong 10 Chemistry, Malinka alaliyadde 12 MSE, and BenCammarata 11 BE.

The same goes for iGEM, established to celebrate and

strengthen the foundation of the scientic community: that

todays advances arrive on yesterdays hard work.

Henning says she found the professional networking, with

the sharing of experiences, an unexpected reward. alaliyadde

agreed. He noted that iGEM gold medals are only awarded when

a team can demonstrate its work has helped another team. They

make it a collaborative community, he says.

The iki-inspired collaborative model is a big part of the

weekends contest, from presentations to a poster-board

discussion of a teams solution.

s a result, Xiong says, the competition gets bigger, better, and

more challenging every year. Every year you have a build-up of

more parts that other teams can use, he says.

Like so many construction bricks and mortar, the BioBricks

stack up with each year. teams success could be directly linked

to a teams past worknext year, it could be thanks to the Cornell

competitors.To save a life, the environment, the world?

s the Cornell students assessed their work sitting around a

table at MITs Stata Center, where high-tech companies handed

out iers for potential future employees, Xiong grinned at the

prospect. Maybe.

The new eill Institute for Cell and Molecular Biology, a state-of-the-artcenter in eill Hall, is a kind of connective tissue for the interdisciplinaryC GEM team.

8/3/2019 CEM Spring 2010

18/36

16 ng 2010

Cornell Engineering Magazine: How is thecollege doing at recruitment?

Mark Spencer:e think the college isdoing a great job. e just came out of a

year in which we saw more applicationsthan the College of Engineering has everseen. e had just over 7,700 applicants

and the enrolling freshmen class of 2013was 37 percent female and 12 percentunderrepresented minorities. ll thosestats are all-time highs for the college.

CEM: How are things looking this year?

MS: e have seen another increase inearly-decision applications. In the last twocycles alone we have seen an increase inearly applications of 33 percent. I dontknow yet where were going to be withregular decision applications, but wereexpecting another increase. e may reach

another all-time high in total applicationsto the college.

(Editors note: As of March 1, 2010,Engineering applications did in fact reachan all new record high of just over 8,400applications!)

CEM: What is the effect on our acceptancerate?

MS: e only have 735 slots each year,so as applications rise, our acceptancerate drops. Since 2005 our acceptancerate has dropped from 38 percent to

22 percent for last year. So, were moreselective, and the quality of our studentscontinues to be excellentfor example,the mean ST score (Critical eading andMath composite) for accepted students isconsistently between 1450 and 1460, outof 1600. lso, for students who have a highschool rank, 96% of them fall in the topdecile of their graduating class. Havingso many applicants gives us even moreoptions for diversifying the class in somany ways.

CEM: Whats behind the increase inapplicants?

MS: There are a lot of factors. e havea great staff and they work very hard.e logged something like 4,000

Engineering dmissions visits last year.nd the Engineering mbassadors, thestudents who give our tours, are greatrepresentatives of the college. ur facultymembers help too by opening up some oftheir classes for visitors to sit in on, or byserving on panels for on-campus eventslike Cornell Days.

nother item that has helped is that thestrategic communications of the collegeand the university as a whole have donea good job representing Cornell to high-school students. The university is denitelya partner in this. From an enhanced ebsite to updated print communications,I believe we have reached out to thesestudents in an informative and positivemanner. e also went to the Common

In the strategic plan laid out in 2005, the college set itsel the goal o becoming a leader in the education o women andunderrepresented minorities. In recent years, the college has passed signifcant milestones on the way to achieving this

goal. We sat down in December with Mark Spencer, director o Engineering Admissions, to fnd out whats behind the

numbers and talk about other recruitment trends.

A conversation with Mark Spencer,Director o Engineering Admissions

By Robert Emro

8/3/2019 CEM Spring 2010

19/36

Cnell engneeng 17

pplication about ve years ago. Someof our increase may have come fromthat change, although not all of it. ndperhaps with this economy, more parentsand students are focusing on studyingsomething with a clear path to a futureprofession where they can make a nice

salary upon graduation.

CEM: What has led to the success inrecruiting women and underrepresentedminorities?

MS: utreach programs that bringprospective students to our beautifulcampus so they can experience the qualityof our programs, faculty, and facilities rst-hand. For example, last fall (2008) about90 percent of the students that attendedeither our omen in Engineering Day orour fall underrepresented minority hostingevents ended up applying. In the springof 2009, with our Prospective Candidateseekend program for admitted femalestudents, 70 percent of those attendeeschose to come to Cornell. These areamaing rates. bviously the efforts ofour Diversity Programs ofce and theinvolvement of faculty and student groups,like the Society of omen Engineers, arecrucial to the success of these on-campusevents.

The Diversity Programs ofce alsohas two summer outreach programs,CIE for women and CTLST forunderrepresented minority students.These programs are stellar and theirefforts in exposing high-schoolsophomores and juniors to engineeringare paying off for us.

Lastly, I have to believe the recent dramaticchanges in our nancial aid packaginghas helped us be more competitive withour peers and attracted more studentstowards Cornell.

CEM: What challenges do you face?

MS: Continuing to build on the success

of this talented and diverse class. Thecolleges goal is to have a total femaleundergraduate population of 35 percent.ight now were at about 31 percentoverall. For underrepresented minorities,the goal is 10 percent and were at 8percent overall. e still have work to do.

nd our competitors are out there tryingto recruit the same students. They arentgoing away.

Its an enormous challenge to select theclass from a group of such accomplishedand talented students. e expect morethan 8,000 applications, and the qualityof our applicants continues to rise. It isextremely difcult to deny admission toso many excellent students who couldundoubtedly make a great contribution tothe Cornell community.

CEM: What do you look for when readingapplications?

MS: cademics is still the biggest decidingfactor. Its important that students havechallenged themselves academically intheir high school and been successful.Beyond that, we look at how well theywould match with Cornell. How do theyspeak about Cornell and engineering?hat do they know about us? Have theydone science or engineering research? rethey leaders? Do they work hard? hatwere their extracurricular activities? Thereis no exact formula here for admission.Each student has a story to tell; whatis that story? For example, a successfulapplicant doesnt necessarily need to haveparticipated in a lot of extracurricularactivitiessome are really committedand good in one or two things, like theyrea licensed pilot or certied airplanemechanic. That is their story.

Frankly, probably 80 percent of theapplicant pool could do the work here.But do they t academically and willthey contribute something to the Cornellcommunity? ne way in which we try

to determine that t is with our Cornellsupplemental essay question. few

years back we changed the questionto be more engineering-specic. erelooking to see if they have an engineeringidea or interest and how they think theycan develop that at Cornell. Its basically

asking, hy engineering? and hyCornell? ere not looking for anythingin particular, but how they are thinkingabout it and whether it makes sense. Itdoesnt necessarily need to be a athanielHawthorne piece of writing, but it is agood way for students to differentiatethemselves.

CEM: What majors are applicants mostinterested in?

MS: Mechanical and electrical engineeringcontinue to be the most popular, but

there is growing interest in elds that arebiology or sustainability related, such aschemical and biomolecular, biomedical,and environmental engineering. Thestudents applying today were alivewhen gas hit $5 a gallon. They sawAnInconvenient Truth. ow they want tomake an impact on the world. ereseeing more altruism expressed now instudents applications. Students wantto develop sustainable water systems indeveloping countries, design the next-generation gas-efcient automobile, andcure disease, to name a few. These are the

people we want to be leading the worldinto the future. They see that engineeringwill not only give them a secure livingbut it will also make the world a betterplace. Theyre going to be difference-makers out there. I can see and feel theexcitement between the lines of theirapplications. This is my sixteenth cycle inthe admissions world and this altruismtrend is something I am seeing currentlymore than any other time in my career. Itsuplifting.

8/3/2019 CEM Spring 2010

20/36

18 ng 2010

educatinginnovatorSKee Feo Pom offe uque etepeeutucto to eee tudet.

Thomas Murray 10 ME (left) got his hands dirty on the Ecoock production line.By Bridget Meeds

8/3/2019 CEM Spring 2010

21/36

Cnell engneeng 19

hen ndrew J. Kessler left Cornell in 1980, B.S. in electrical

engineering in hand, he was at the beginning of an untraditional

but very successful career. For the rst 10 years, he learned the

ropes at big corporationsT&T, Paineebber, and Morgan

Stanley. But then he struck out on his own, hitting the sweet

spot of the 90s tech boom with elocity Capital Management, a

venture capital fund. ow he writes books and frequently appears

in the media commenting on tech entrepreneurship. Its been a

great ride, but he had to learn everything about running his ownbusiness on his own.

Kessler has decided to train a new generation of entrepreneurial

engineers, but in a way that gets them out of the gate quicker,

by providing them signicant support to learn the business side

of entrepreneurship as undergraduates. In 2009, he gave the

College of Engineering $2 million to fully fund the rst ve years

of the Kessler Fellows Program. Its an innovative combination of

coursework and internships that teaches 10 juniors each year the

know-how to go out and do it on their own.

ndy wanted to develop a comprehensive program that had

to do with education and experience surrounding what he has

termed the business wrapper of entrepreneurism, says program

director Tracey Brant. Each student attends a course in the spring

of their junior year about the essentials of entrepreneurism, in

which they hear war stories from alumni entrepreneurs and

learn real business skills. They spend the summer interning

with a start-up. nd in the fall of their senior year, they present

a symposium series, sharing what they learned with the

engineering community. In return, they receive a salary for their

internship, travel expenses, a cash prie, and, if needed, a portion

of their senior year nancial aid package is paid by the program.

In summer 2009, Kessler Fellows did traditional engineeringmeasured nanoparticles, tested the re resistance of building

materials, wrote computer codebut those same fellows also did

market research, wrote business plans, and researched patents.

Not enough hands

Everything I have is pretty darn important and I dont have

enough hands, says Ken ang 77 ChE, president of Hybrid

Silica Technologies Inc. e actually need help more than larger

companies.

Peter Bai 10 MSE soon discovered this with his internship at

the start-up, born in the lab of lrich iesner, Spencer T. lin

Professor of Materials Science and Engineering. HST develops

uorescent core-shell nanoparticles with biomedical, energy, and

electronics applications at its labs in Ithaca, .., and Cambridge,Mass.

t HST, Peter measured nanoparticles and analyed data in the

lab. He made training videos on lab techniques and created a

private ouTube account to host them. He researched patents.

He helped plan the companys expansion to Cambridge, learning

about permits and vetting potential lab spaces. nd in his free

moments he wrote a 25-page business plan about a spin-off

product that HST hopes to market, one that will help identify

conditions to which diabetics are prone.

It was a great learning experience, says Bai. It was exactly

what I wanted to be able to do, to experience many aspects of theentrepreneurial environment.

The Kessler Fellows Program is absolutely brilliant, says

iesner. It provides the intellectual environment that

particularly talented students thrive on. e need to not only

educate the best technical engineers, but also individuals that

are ready and hungry to bring the most promising engineering

solutions to the market. That is exactly what the Kessler Fellows

Program helps to do.

Peters very versatile, says ang. In a start-up, you need

people who can help pack a truck one day, and another day, do the

highest level thinking on the IP.

it Was a great learning experience. it

Was exactly What i Wanted to be able

to do, to experience many aspects of

the entrepreneurial environment.

peter bai 10 mse

ProvIDED

Peter Bai 10 MSEdid everythingfrom measuringnanoparticles to

writing a businessplan at Hybrid SilicaTechnologies.

8/3/2019 CEM Spring 2010

22/36

20 ng 2010

ProvIDED

ProvIDED

Scott Purdy 10 CS

NumentaRedwood City, Cali.

Mike Ryan 10 EPV.i. LabsWaltham, Mass.

Joey Zwicker 10 MEAN.Y.botsMountain View, Cali.

Casey Worthington 10 ECEMarketceteraSan Francisco, Cali.

Je Will 10 ChEThe Water InitiativeNew York, N.Y.

Beautiful engineering

Thevaki Thambirajah 97 (pplied Economics and Management),

who owns Thevi Cosmetics and hosted Kessler Fellow my Chen

10 E, agrees. e have so much work to do and we have

limited resources! Thambirajah says.

Thevi Cosmetics in ew ork, .., is a high-end, boldly colored,

all-natural line inspired by the Indian lifestyle and based on the

yurvedic philosophy. How did an engineer end up interning at a

cosmetics company? my is an unusual engineer, one who openly

celebrates her girly side. She found a way to mesh her passion

for cosmetics with business development through the Kessler

program.

s soon as Chen arrived, she dove enthusiastically into doing

hands-on market research at department store counters,

planning and executing a social media marketing plan including

a splashy blog, and creating the mood boards that demonstrate

the feel of the brand for Thambirajahs investor presentations.

In the process, Chen helped change the very direction of the

company from being a brand targeted to Indian women to an

Indian lifestyle-inspired brand that appeals to more consumers.Initially, before my came on board, we were very focused

toward the minority market, says Thambirajah. But we found

that retailers as well as consumers want to have a broader reach.

So through my and mys research, we found that the Indian-

inspired concept had a wider appeal.

Thomas Murray 10 ME worked on technical development of Ecoock inthe lab at Serious Materials.

my Chen 10 E helped Thevaki Thambirajah 97 reorient TheviCosmetics toward a broader market.

Peter Bai 10 MSE

Hybrid Silica TechnologiesIthaca, N.Y.

Amy Chen 10 OREThevi CosmeticsNew York, N.Y.

Bhaskar Garg 10 MEPalo Alto Research CenterPalo Alto, Cali.

Tucker Moat 10 MEBug LabsNew York, N.Y.

Thomas Murray 10 MESerious MaterialsSunnyvale, Cali.

2009 Kessler Fellows

8/3/2019 CEM Spring 2010

23/36

Cnell engneeng 21

I felt really honored, says Chen, because Im still in college,

and she trusted me enough to really tell her my opinions aboutthe direction she should take her company. It was a really great

experience and I got to accomplish a lot of things that I honestly

never dreamed Id be able to do just as an intern.

A new field needs new ideas

Entrepreneurs benet from interns because interns dont have

any preconceptions about how things are done, says Dan Boss,

vice president of engineering for Serious Materials in Sunnyvale,

Calif. Serious Materials is riding the eitgeist of green building

by manufacturing building products that help reduce the

embodied energy of a buildingincluding energy used for raw

material extraction, transport, manufacture, construction, and

deconstruction as well the energy usage in its operation. They

offer highly-efcient windows and Ecoock, the low embodied

energy drywall on which Thomas Murray 10 ME focused.

Interns come in open-minded, continues Boss. hen youre

an entrepreneur, thats one of the key aspects that makes you

successful, to look at problems from a different perspective.

Murray says he was really pleased to offer his different

perspective to Serious Materials during his internship. During

the summer, says Boss, Tom was focused on Ecoock. e had

him do both some technical development for us, looking at somechemistry of the system, but also looking at the value proposition

of Ecoock versus competitive products. He had the chance to get

his hands dirty in the lab and work on the production line as well

as look at how Ecoock ts in, particularly for LEED construction

in the commercial space.

Traditional drywall has been made in the same way for more

than 100 years using gypsum, a limited natural resource, with

an energy-consuming, carbon dioxide-producing drying process.

Ecoock is air-dried, using 80 percent less energy; is made from

80 percent recycled waste from steel and cement plants; can end

its life cycle as a soil amendment; resists mold better; emits nomercury; and looks great. Builders earn LEED credits when they

use it.

ne big thing that I learned from the Essentials of

Entrepreneurism class and then saw in practice was that bringing

a product to market is one-third engineering, says Murray. The

other two-thirds are business and marketing. orking for a larger

company, you are just going to be able to see your engineering

part of the process. But working in a start-up, youll see the entireprocess.

Thomas is developing a very valuable perspective in that hes

looking at both the technical problem and the business aspect

of it, says Boss, which I think it is a great focus for the Kessler

Program, because thats often missing. lot of the technical

internships focus just on the technology and it is the synergy

of the business and the technology that makes something

successful. If you dont have both aspects, you only have half the

answer.

The entrepreneurs of the futureEach Kessler Fellow has taken the skills and opportunities of

the program and used them to plan the future. Bai is applying

to graduate programs with good technology transfer support,

envisioning developing his own company before he nishes

his Ph.D. Chen plans to enter the cosmetics industry to learn

its intricacies and hopes to eventually open her own company.

Murray is spending the spring having conversations with

members of the Kessler advisory board, bouncing around ideas

about the sustainable building business hed like to start.

This program really saved me in a lot of ways, says Chen,

who took a long time to nd her niche in engineering school. Iwas grasping for anything to help me with my career, and then

I found this program and it opened up this whole new world for

me. Honestly, without ndy Kessler and the help hes given me, I

dont know where I would be right now.

In the lab, a lot of people have very good ideas, says Bai. But a

lot of those ideas are being buried in notebooks because not a lot

of people are taking the next step to develop them into products

that have commercial value. It is important to bring these ideas

into the marketplace because it not only benets the people

who invent and sell these products; it also generates value for

everybody.The Kessler Fellows is an awesome experience, says Murray.

Theres really no other time in your life when youre going to be

able to just go and pick the start-up you want to work for and go

see that culture and have it be a wonderful learning experience.

this program really saved me in a lot of Ways. i Was grasping for anything to

help me With my career, and then i found this program and it opened up this

Whole neW World for me.

amy chen 10 ore

8/3/2019 CEM Spring 2010

24/36

22 ng 2010

In the nal act of the 1956 lm Forbidden Planet, Robby the

Robot is ordered by his creator, Dr. Morbius (Walter Pidgeon)

to destroy an invisible monster descending upon the doctors

laboratory. As the creature hammers the steel walls that shield

the lab, Robby shakes, rocks, heats up, and shuts down. Te

robots inaction is not insubordination but a built-in safeguard

that prevents Robby from harming his master: although

Morbius has not yet admitted it, the invisible beast outside is an

extension of himself, which Robby already knows.

Forbidden Planetis one of the earliest lms to exploresafeguards hardwired into robots built to serve man. Its a

concept in full bloom in the robot novels by Isaac Asimov, in

which robots are programmed with three laws intended to

protect mankind from runaway robots. For the last half-

century, popular entertainment has explored the possibilities

and consequences of robotics and articial intelligence, from

the sublimely sinister in Stanley Kubricks 2001: A Space

Odyssey and its cunning self-aware computer HAL 9000, to the

emotional in Steven SpielbergsA.I.

But as science ction runs amok with tales of articial

intelligence leveraged for evil and greed, scientic advances in

Killer robThe promise and perils of artificial intelligence

Built-in safeguards prevented obby the obot, from the 1956 lmForbidden Planet, from obeying Dr. Morbius when he ordered the robot todestroy an invisible menace that was an extension of himself.

CorBIS

By Michael Gillis

8/3/2019 CEM Spring 2010

25/36

Cnell engneeng 23

These systems are really starting

to be part of our society. And the

question is, what rules should they

obey? What are the risks?

Bart Selman

CBIS

the real world have quietly and signicantly expanded the

boundaries of AI, and it looks a lot dierent from pulp ction.

Many of these advancesand questionsare evident at the

intersection of AI and robotics. Semi-autonomous machines are

already at work in factories and airplanes, and on the battleeld,

where unmanned drones fulll missions over Iraq and

Afghanistan. As that technology advances and migrates to more

systems, and as those systems become more autonomous and

sidestep human control, the possibilities are spectacularand

potentially frightening.We have paid very little attention to the societal impacts or

the ethical issues, because up until now, these systems were not

very feasible, says Bart Selman, Cornell professor of computer

science. Selman co-chairs a yearlong study into AI and society

by the Association for the Advancement of Articial

Intelligence. Te study comes at a critical time, Selman says,

because advances over the next 15 to 20 years will bridge

science ction with science fact.

Tese systems are really starting to be part of our society, he

says. And the question is, what rules should they obey? What

are the risks?

otSHAL 9000, the conficted computer in2001: A Space Odyssey, was programmed to deceivewith disastrous results or its crew.

Bart Selman, Cornell professor of computer science and co-chair of a yearlong study into I

and society by the ssociation for the dvancement of rticial Intelligence.

The AI explosion

Te mystery of intelligence remains a tough nut to crack, but

computer scientists have made important advances in several

elds, including machine learning, which teaches computers

how to behave based on data and complex algorithms.

Progress may be slower than expected decades ago, but Eric

Horvitz, president of AAAI and a senior articial intelligence

researcher at Microso, says the last 25 years have yielded

insight into key aspects of intelligence, including perception,representation, learning, and reasoning.

In the last decade, there has been an acceleration in our

understanding, our prowess in both principles and applications

on multiple fronts based on a coalescence of several factors,

Horvitz says.

Banks lean on AI to manage investments, hospitals use AI to

diagnose illness and manage work schedules, and the

government leverages AI in the form of speech and gesture

recognition and data mining to help curb cybercrime and track

down terror suspects.

One ubiquitous AI application is the aviation autopilot, used

8/3/2019 CEM Spring 2010

26/36

Tsuhan Chenswork puts him at th

xus o svra discipis with o

thi i commovisua data. His

projcts ivov vrythi rom capturi

ad haci th sia st rom a

diita camra, to aayzi it or pattrrcoitio, to usi it to rdr diita

imas o what th camra did ot s.

projct Chs roup is worki o

coud provid surviac with wr

camras ad astr ractio tims. Th

oa is to proram roups o tworkd

robocams to patro buidis ad

historica sits. imp ima aaysis ts

thm whr thrs movmt, but it taks

mor sophisticatd aorithms to idtiy

suspicious activity. W d to kow

what our tart is, says Ch, dirctor

o th choo o ectrica ad Computreiri. W wat thm to kow

wh somthi is itrsti ouh

to rcord, ad scod, w wat thm to

work tothr to t th bst ima.

a coup o th robocams brak dow,

or xamp, th othrs d to adjust to

covr th aps. each robocam has its ow

computr, so dcidi who os whr is

a joit dcisio btw para prorams.

Th roup dcisio is mad i a vry

distributd mar, says Ch. Just ik

humas.

Hadas Kress-Gazitwats to kow

how to mak robots do what thy ar

supposd to.

c a robot udrstads th mai o

a commad iv i atura aua, it

must b trasatd ito motor ad vocity

commads. Thats whr Krss-gazit

coms i. You wat to say, arch th

rooms. You dot wat to say, Mov fv

mtrs orward. Tur 30 drs to th

riht, ad this ik that, sh xpais.

You wat to kp it at th hih v.

To b usu, a robot must b riab.

wat to hav som kid o uarats or

th robot bhavior, bcaus aythi ca

b hackd tothr, sh says. wat a

systmatic way o oi rom a hih-v

dscriptio to a corrct bhavior. Ad i

cat, wat to kow why caot.

Thrs o shorta o ways to t a

robot to simpy o rom poit A to poit

B. What Krss-gazit is itrstd i ar

tasks that rquir robots to ract to thir

viromt. Thr is ot a ot o rsarch

out thr to da with i wat to say, K.

wat you to o throuh th rooms.

you s my dauhtr, just stop thr ad

ca m, sh says. Th bhavior o th

robot is oi to cha basd o i my

dauhtr is aywhr i thr or i sh is

movi aroud or i shs just ot thr.

shs ot thr, th robot is just oi

to cotiu sarchi. Th bhavior is

ractiv.

For xamp, sh cotius, i t th

robot, go sarch a th rooms ad mak

sur you t ito vry si room, ad

thr ar doors. wat th robot to t

m, list, cat op doors. This is th

raso caot do what you askd m to

do.

o ar, Krss-gazit has ot hr robots

to t hr wh thy caot prorm a

commad, but ot why. Thr ar may

othr qustios sh woud aso ik to

aswr. iht ow m assumi that i

th robot is sarchi or my dauhtr ad

my dauhtr is thr, it wi s hr, sh

says. it dost s hr at a bcauso its poit o viw ad ts m shs ot

hr, what dos that ma? s that corrct

bhavior? s that icorrct bhavior?

Thrs a ot o ucrtaity i th ssi

that has to b icorporatd.

Ashutosh Saxena tachs robots to

oprat autoomousy i w, ucrtai

viromts.

Most xisti robots ca b scriptd

to prorm difcut tasks i hihy

costraid, kow viromts.

The New Roboticists

More than 20 Cornell aculty members are solving problems in robotics and artifcial intelligence. Their research

spans various subareas, including perception, control, learning, planning, and human-robot interaction. They

work with aerial robots, home and ofce assistant robots, autonomous cars, humanoids, evolutionary robots,

legged robots, snake robots, and more. Here are three recent additions to this group.

Tsuhan Chen Hadas Kress-Gazit

By Robert Emro

24 ng 2010

8/3/2019 CEM Spring 2010

27/36

Cnell engneeng 25

extensively on commercial ights. Te technologys benets are

obvious, but it also provides an example of potential dangers in

an automated society.

en years ago you could just turn o the system, Selman

says. However, he explains that switching o autopilot may not

be the right choice today. He points to a recent incident in New

York where a pilot turned o his planes automated controls and

crashed. Its very hard for a pilot to take over in a dicultsituation. Because they have been ying autopilot so much, they

may not actually have the training anymore to recover from

dicult situations.

Tat growing dependence on automated technologies is why

Selman and the AAAI are taking a hard look at the societal

eects of AI.

AI research is at the front edge of a larger computational

revolution in our midsta technical revolution that has been

introducing new kinds of tools, automation, services, and new

access to information communication, reads the interim report

of the AAAI Presidential Panel on Long-erm AI Futures.

Ghost rider

Some of the most obvious advances in AI are found in the

eld of robotics, in part, because its easier for people to see

these machines as physical beings, Selman says. Robots are

already widely used in manufacturing and the medical

profession, and now robotics stands at the dawn of a new age.

In 2005, eam Cornell participated in the Defense Advanced

Research Projects Agencys second Grand Challenge, one of a

series of contests launched by the U.S. military in 2002 that

oers millions of dollars to groups that build vehicles that cannavigate themselves across a nish line.

In 2007, Cornell was one of six teams to nish the DARPA

Urban Challenge at a former military base in California. Its

driverless Chevy ahoe, dubbed Skynet aer the AI in Te

erminatormovies, traveled 55 miles across a variety of terrain,

obeying trac lights, merging with trac, and parking itself.

Selman, who sits on the advisory board for the Cornell team,

says the contest and its achievements are proof that AI is

grown up.

Tose accomplishments also help paint a more vivid picture of

the potential societal benets of driverless vehicles.Such robotic systems will rst come to the fore as collision

avoidance systems that will sense dangerous situations and take

over driving briey, Horvitz says. Even early versions of

robotic automobile systems promise to reduce trac injuries.

We have had over 40,000 people killed each year in the United

Statesand I believe that most of these deaths will one day be

prevented by robotic vehicles.

Tat kind of technology will likely be commonplace in 20

years, says Selman, which is why the AAAI is raising the

questions of responsibilityand quelling fearas robots rise.

That works most o th tim, but it dost appy to actuaviromts whr thr is ucrtaitity ad thats a bi

probm, says axa. o my oas is to ab robots

to prciv th viromt, v i its compty w, so that

thy ca oprat i that viromt. t has ot oy to ir

th 3-D structur o th viromt ad rcoiz objcts,

but aso to fur out what to do with th objcts. A robot is

ot just a passiv obsrvr; it is mat to prorm dsird

tasks by acti upo th viromt.

axa uss machi ari aorithms to tach his robots.

W dot hav to proram th computr xpicity; you t

thm ar rom th data, h says. Just ik wh you ar

traii a do. By showi robots samp probms ad

thir corrct aswrs, axa ca trai robots to do a sortso tasks. At taord, whr h studid or his h.D., axa

usd machi ari to sov svra robotics probms,

icudi how to crat a thr-dimsioa mod rom a

two-dimsioa ima.

W showd th computr about 500 imas ad th corrct

3-D mods or th imas, says axa. o it ard th

uctioa mappi rom th ima to th dpths. Th wh

iv a w ima it coud mak a 3-D mod.

With this tchooy, axa turd a rmot cotro car ito

a robot that coud driv its throuh uamiiar surroudis

u o obstacs, ik a orst. ts a sma robot, so it caot

hav bi ssors o it, but it ca us camra imas to fd

out whr it shoud o, h says. Ad it coud ray driv astavoidi th obstacs, v os that wr ot s bor.

axa is ow tachi robotic hicoptrs how to y without

th aid o positioi systms. ts ray hard to mak

aria robots y i g-did viromts, or i cuttrd

viromts, h says. Thrs o robot that ca y isid a

buidi i costraid spacs.

axa has traid a robot to uoad itms rom a dishwashr

ad wats to buid a ra purpos hom ad ofc

assistat robot that ca do houshod chors ik tidy up a

room, tch or divr itms, or prpar simp kitch mas.

Ashutosh Saxena

8/3/2019 CEM Spring 2010

28/36

Bad robot

Te AAAIs interim report takes unabashed aim at science

ction.

It points out that much of the lay perception of robots and

articial intelligence is fueled by pulp ction visions of the

technological singularity, when humanity creates an

intelligence greater than its own that can then design even

smarter machines. Te result is either a world overrun by killer

machines, as in Te erminatormovies, or a utopian future

where AI solves all our problems.

Te panel of experts was, overall, skeptical of the radical

views expressed by futurists and science-ction authors, the

report reads.

Te interim report specically mentions science ction author

Isaac Asimov and his robot novels. Robots and popular culture

are much informed by Asimovs writings, the ideas of which

populate an ever-expanding mythology in movies and books.

Asimovs laws of robotics are well known throughout thescience-ction community, and many readers believe they are

sensible rules to follow: 1. A robot may not injure a human

being or, through inaction, allow a human being to come to

harm. 2. A robot must obey any orders given to it by human

beings, except where such orders would conict with the First

Law. 3. A robot must protect its own