Unit0

WhatisArduino

Referredtohttps://learn.sparkfun.com/tutorials/whatisanarduino
Arduinoisanopensourcesingleboardmicrocontrollerusedforelectronicsprojectssuitableforboth
beginnersandadvancedusers.TheyaregodsendforDIYenthusiastsandtinkerersduetoitsflexibility
andunlimitedapplications.ItoriginatedfromItalybackin2005.ArduinoconsistsofanAtmel8bit
processer.TheArduinoUnoR3weuseinthiscoursefeaturestheATmega16U2, thereby offering
greater transfer rates and memory over its predecessors.

WheretofindResources

ArduinoWikipagehttp://en.wikipedia.org/wiki/Arduino
Arduinomainwebsitehttp://arduino.cc/
https://www.sparkfun.com/products/retired/11227

ExaminingtheSparkFunInventorsKit

Justlikeopeninganythingnewelectronicforthetime,Iwasreallyexcitedtofinallycrackopenthebox,
butitwasmorelikeboxofchocolates.Manylittlecomponents,someofwhichIrecognizelikethe
arduinouno,breadboard,motors,etc.

InstallingtheArduinodevelopmentenvironment

FollowedtheSIKinstructionsforinstallation.Usedthewindowsinstaller,prettystraightforward.ThenI
connectedthearduinounotomycomputerusingtheusbcable,andthedriverswereinstalled
automatically.NextIlaunchdthearduinoIDE.TheIDElooksamazing,verysleekandmodern.Butthisis
notsurprisingastheirwebsiteandoverallbrandinghaveaverysleek,simpleandmodernfeeltothem.
TheSIKguideandtheUnoboardalsodontwell,goodlookandbuildqualityrespectively.IntheIDE,my
boardandCOMportwerealreadyattherightselection.
Referredtohttps://learn.sparkfun.com/tutorials/installingarduino

Writingandexecutingyourfirstprogram

SIKCircuit1

MYfirstprogram!Ifollowedtheguide,itwasstraightforward,learnedaboutpinMode,digitalWriteand
delayfunctions.TheprogramransuccessfullywithLEDblinkingonandoffasexpected.ThenIchanged
thepinvalueinpinModefunctionandrantheprogramagain.Itworksafterofcoursemovingthejumper
cable.IthenchangedvaluesindelayfunctiontochangerateofLEDblinking.Makingadelay0makes
theledasexpectednotblinkatall.

VIDEOhttps://www.youtube.com/watch?v=03teMswFH8

Unit1

Unit1WorkbookExercisesRoboticComponents

Exercise1

1.ListSparkfunssensors

Tempsensor
FlexSensor
Photocell
Softpotentiometer

2.DefineSparkfunssensorspace. For each of the sensors, identify what it perceives and the range of
possible sensor readings.

Temp Sensor: The one in the sparkfun kit is a low voltage centigrade temperature sensor that can
measure temperature over the -40C to +125C with an accuracy of +or- 2C

Flex Sensor: Its 2.2 in length and measures flex as its bent (its resistance changes).Not meant to flex at
its base.

Photocell: Can sense ambient light as resistance changes based on exposure to light. Light resistance
approx.. 1K ohm, dark resistance approx. 10K ohm

Soft Potentiometer: Can sense the relative position on a strip. Resistance changes from 100Ohms to
10000Ohms as we press down on the strip

3. Now imagine a sonar sensor has been added to Sparkfun. What is the new sensor space?

With a sonar sensor, the sensor space would include the ability to detect objects by sound (pings) , the
result of such detection would be a distance value.

Exercise 2

1. Describe the difference between effectors and actuators.

An effector is the physical part of a robot that can affect the environment, such as arms, legs, etc.
Whereas, an actuator is the mechanism by which the effector is able to perform these operations.

2. List each of the Sparkfuns actuators and effectors.

Effectors:
Red, Yellow, and RGB LEDs
Piezo Buzzer

Actuators:
DC Motor
Servo



3. List the degrees of freedom for the Sparkfun.
Unlike the Create, for which these questions were intended for, the Sparkfun does not move and so I
would say it has 0 Degrees of Freedom.

Chapter1FoodForThought

Whatelsecanyoudofromafar,bymeansofteleoperation?Youcantalk,write,andsee,asin
telephone,telegraph,andtelevision.Thereismore.Canyouthinkofit?

Youcancontrolmanythingsremotely.InfactIamremotelytypingthismessageonatabletinascreen
sharingappconnectedtoaPC.YoucanremotelycontrolRCcars,UAVs.Youcanremotelylistento
musicbymeansofBluetooth.Youcanremotelycontrolamousepointerwitha2.4ghzwirelessmouse.

Isthermostatarobot?
Iactuallydidntrealizethis,butaccordingtothedefinitionofarobot,athermostatisindeedarobot.
Thisisbecauseitisatangibleobjectintherealworldthatcansensetemperatureandreactaccordingly
toperformitsobjectives.

Istoasterarobot?
Bydefinition,atoasterisnotarobot,becausethoughtitexistsintherealworld,itdoesnotsenseor
reactonitsown.Itsamechanicaldeviceontimerthatrequiresphysicaloperationtomeetits
objectives.

Someintelligentprograms,alsocalledsoftwareagents,suchasWebCrawlers,arecalledsoftbots.Are
theyrobots?
No,becausetheydontexistphysicallyintherealworld,theyarenottangible.

IsHAL,fromthemovie2001,theSpaceOdyssey,arobot?
MostdefinitelyYES!HALcouldsenseandreactwithhisenvironmentandperformvarioustasksinthe
realworldonhisown.

Chapter2FoodForThought

Howimportantisitforrobotstobeinspiredbybiologicalsystems?...Doesitmatterwhatkindofrobot
youarebuilding?Doesitmatterifitisgoingtointeractwithpeople?

Ithinkitisveryimportantforrobotstobeinspiredbybiologicalsystems,mostofthetimeirrelevantof
whethertheyinteractwithpeopleorthekindofrobotyouarebuilding.Firstly,becauserobotsbasically
performtasksthathumansdo.Nowwecantrytodeviatearobotscapabilitiesfrommimickinga
humans,likeairplanes,etc.Butarobotwouldstilloperatewithinournature.Andbiologicalsystemsin
ourworldhaveevolvedtoworkinthisnature.Sowhathappensis,robotswouldatsomelevelinour
worldalwaysatsomelevelgetinspiredbysomebiologicalsystem.Soforhelicoptersitwouldstillbe
wings,andperhapsflightofsomeinsects.Nowifwebuildrobotsinouterspace,wherebiological
systemsmightnotreign,itmightstarttobedifferentfrombiologicalsystems.Forstarters,dealingwith
antigravity.Forinstance,anobjectthatcanfloatinspacewiththrustersinvariousdirections,isnot
reallyinspiredbyanyorganismsflightinnature(ordeepsea),sincegravityplaysaroledownhere.So
thiswouldbetheexception.

Chapter3FoodForThought

1.Whatdoyouthinkismoredifficult,manipulationormobility?Thinkofhowthoseabilitiesdevelop,in
children,babies,andadults.

Babiesfirstdeveloptheirbrainsandtorso.Theythendevelopgrossmotorskills,wheretheycanmove
theirhandsandlegsroughlybutnotinfullcontrol.Afterthisstagetheystartcrawlingandeventually
walkslowly.Duringthetimetheylearntowalkwell,theyalsodevelopfinemotorskills.
Soitishardtoanswerthisquestiondefinitively.Iwouldsay,basiccrudemanipulationiseasierthan
mobility.Butsomeprecisemanipulationisharderthanbasicmobility.Andadvancedmobilitylike
specialmaneuvers,etc.areroughlyjustasdifficultassomeadvancedmanipulation.

2.Howlargedoyouthinkyoursensorspaceis?

Wellwehaveour5senses.Wecanfeeltouchliketextures,temperatures,etc.Wecansee:wehave
peripheral20/20visionat24fps,wecanseevisiblelight.Wecansmellbymeansofourolfactorynerves,
sotheyhavealimitation(worsethandogsforexample).Wecanhearhear,between20and20,000Hz
frequencies,atlimiteddecibels.
Nowitseemswecansensemuchmorethanthis,butthatsonlybecauseofwhatweashumanscan
dofromthesesenseinputs.Forinstancewesmellsomething,andthatinstantlytriggerspastmemories,
itcouldbeatraumaticevent,andsowefeelwesensedsomethingmore.Ithinkthisisalsohowsome
peopleclaimsixthsense.Theystillhave5senses,itisjusttheirbrainsunderlyingprocessestagged
alongwiththeir5sensesthatcreatesthatillusion.Anotherexample,isthefeelingoflove,westillhave
our5senses,itsjustourbrainprocessing(releasingchemicals,etc)intandemwithoursensesthat
makesusfeellikewesensemuchmore.
Sooursensorspaceislimited,butisseeminglyvastgiventheprocessingbyourbrains.

3.Canyouthinkofinformationinlifethatisobservable,partiallyobservable,hidden?

Observableinformationincludesthingsoursensescanpickup,likeobjects,sounds,etc.Hidden
informationincludesthingslikeemotionalfeelings.Partiallyobservableinformationcouldbethingslike
Iexplainedinthepreviousquestion,whereyouobservesomebasicinformationbutourbrainperceives
morebyprocessingthatinformation.Soforexample,instincts,wecanobservesomethingsubtlebut
dangerous,andthenourbraintriggerspastmemories,andsowecanbyinstinctrunawayfor
example.Buthappenedwasweonlypartiallyobservedsomething.

SIKCircuit2

Thiscircuitissameascircuit1exceptthereisapotentiometeradded.Thispotentiometercanoutputa
voltageanywherebetween0and5volts.Thisiscontrolledphysicallybytheuserturningtheknob.My
potentiometerhadapinharshlybent,anditisflimsy,feelslikeitcanbreakanytime.Butotherthanthat
Circuitwasjustastrivialtosetupasthelasttime.Rotatingthepotentiometer,theLEDblinksfroma
slowintervaltofast.Thisisaccomplishedbythedelayfunctiontakinginthevaluebeingoutputted
(analog)bythepotentiometer.

VIDEOhttps://www.youtube.com/watch?v=Vsl8_5eURK0

SIKCircuit3

WhenIfirstputtogetherhastilyImissedtheresistors,butnoticedmyerrorandimmediatelyputthem
in,butthennothingworked.IthenrealizedIputtheGNDwireonthewrongsideofthebreadboard.
AfterfixingthatItworked.Followingtheguide,everythingworkedwithoutahiccup.Ijusthadtobe
extracarefulingettingthepinsinrightbyseparatingthem,andknowingtheshortoneisRed(sequence
isRed,GND,green,blue).Theprogramissimilartothepreviouscircuitsexceptitturnsonthevarious
Red,Green,BlueLEDsindifferentcombinationstooutputthedesiredcolor.

VIDEOhttps://www.youtube.com/watch?v=dnH7MVEjr1Y

SIKCircuit4

Thisprogramwasalsosimilartopreviouscircuits,exceptwith8LEDs.Assemblingwasjusteasythough.
Exceptputtingallthejumpercablesintodigitaloutputpins29wasannoying.Thepinsweretooclose
togetherandsothejumpercablesdidnothaveenoughroomnexttoeachother.Isomehowmanaged
tojamthemin.IusedalternatingyellowandredLEDsasopposedtoallsamecolorbecauseIplannedto
dosomechangestotheprogramforfun.Ifirstranthecircuitasitwasintendedanditwentsmoothly.It
wasjustaforloopintermsofcode.ThenIchangedthecodeinthelooptoturnonallevenLEDSonat
once,thenoff,andthenalloddatonce,thenoff.Itwassimple,butoutputwasasexpected,allyellow
LEDsturnedonatonceandthenallRedLEDsturnedonatoncenext,andsoon.
VIDEOhttps://www.youtube.com/watch?v=UIoPhlTdA4Q

SIKCircuit5
Finally,somethingotherthanjustLEDs.Iassembledthecircuitaccordingtotheguide.Seemedsimple
enough,exceptoneofthepushbuttonsIgothadallitspinsbent.Ihadtoalignthemcorrectlytofitinto
thebreadboard.AfterfinishingthecircuitIrantheprogramandnothingworked.Icheckedthephysical
connectionsacrossmyboard.Ihad2mistakes,Ididnothave1ofthewiresalignedcorrectlywitha
pushbuttonandIhadmyLEDwireinthewrongpin.Thecodeworkslikethis:IfIpusheitherbuttonit
connectstogroundandtheLEDwilllightup.IpushboththenbyBooleanlogictheLEDwillturnoff.It
worksasexpected.
VIDEOhttps://www.youtube.com/watch?v=5VfuqJ6efOw

SIKCircuit6
Iassembledthecircuitbytheguide.Assemblywassmooth.Thecircuitisstraightforward,exceptthe
onlythingtonoteisthatavoltagedividerisrequired.Theoutputofthiswillbereadasananaloginput.
Analoginputrangefrom0to1023butwriteonlyrangesover0to255sotheinputismappedtooutput
rangeof0to255.Theprogramranasexpected.Aswritteninprogramdocumentation,thelightdidnt
completelyturnoff,soIplayedaroundwiththerange0to1023.Ieventuallypicked400700andit
workswellenough.TheLEDturnsoffwhenlightisdetectedbythephotocell.Ialsoreversedtheeffect
bywritingtotheanalogpin:255thelightlevel.SonowtheLEDisonwhenitlightissensed.
VIDEOhttps://www.youtube.com/watch?v=80_jkZpa3EM

SIKCircuit7
Thiscircuitwasparticularlyeasytoassemble,onlyonepartneededwithnoresistors,justthetemp
sensorandwires.Hadtolearntoidentifythepinsforthetempsensor.Withthesensorflatfacingyou
(pinsdown)theleftmostpinis5v,andrightmostisGND.Ihookedupthewirescorrectly.Finallywriting
codeforoutputtingdataonscreen(serialportwindow).Restofthecodewasstraightforward.Execute
programandgetfollowingresults
voltage:2.89degC:238.57degF:461.43
*Facepalmtohead*
Evenafterknowingwhateachofthe3pinsonthetempsensorare,Imixedupthewirespositiveand
negativeonthebreadboard.FirstIthoughtthereadingwasoffandItouchedthesensoranditreally
washot.Irantitagainafterfixingthewires.Andthereadsweregoodnow,prettyconsistent.
voltage:0.72degC:22.27degF:72.08
voltage:0.72degC:22.27degF:72.08
voltage:0.72degC:22.27degF:72.08
voltage:0.73degC:22.75degF:72.96
voltage:0.72degC:22.27degF:72.08

Voltagefluctuatedupto.73anddownto.71.Thisisrelativelyminoranditsprettystableactually.I
touchedthesensorandthetemperaturespiked,butnothingerratic.Soitwasallnormal.
voltage:0.73degC:23.24degF:73.84
voltage:0.73degC:23.24degF:73.84
voltage:0.73degC:23.24degF:73.84
VIDEOhttps://www.youtube.com/watch?v=03teMswFH8

MeetingLearningObjectives
1)definethetermrobotandprovideabriefhistoryofrobotsandrobotics.
Arobotisanautonomousphysicalsystemintherealworldthatcansensethingsandservesome
purpose.
2)listthecomponentsofatypicalrobot.
Followingrobotcomponentsrelatestoahumanbeings
Bodytoexistinthephysicalworldanddothings
Effectorsandactuatorstodoperformactions
Sensorstosenseitsenvironment
Controllertothinkandactonitsown

Unit2
Unit2WorkbookExercisesLocomotion
ComparewhattheyaredoingwiththeiRoombatorobotkitslikeLegoMindstorms.Howdothey
compare?
TheLegoMindstormskitlookslikeitsgearedmoretowardsayoungeraudienceforasfunhobbyisttoy.
Thoughitcanmostdefinitelyserveotherpurposes,canbeusedforrapidprototypingforsmallprojects
thatcanuseitsarrayofsensors,wheels,tracks,etc.TheiRoombakitlooksmoreprofessionaland
comesfullyassembled.Itcouldbeagoodforresearchapplications.Itisjustarobotbasethatcanadd
moresensorsandhasacargobayformountingotherthings.
HowdothemotorsusedintheiRoombacomparetothevariousmotorsinyourInventor'skit?
TheiRoombahas2motorsattachedtoitsdrivewheels.Icouldnotfindspecsonthesemotors.Iwould
assume,forthesizeoftheiRoombaandthecargoitcancarry,itsmotorsaremorepowerfulthanthe
Skparkfuns.
DiscusshowyoumightbuildarobotliketheiRoombausingyourInventor'skit,includingandadditional
partsyoumighthavetoacquire.
Wow,sincetheiRoombaisacompleteunit,fullyassembled,Iwouldsaythattobuildsimilarlywith
Sparkfun,wewouldneed,wheels,gears,chassis,anadditionalmotor,motorshield.Wewouldalso
needahostofsensors,likedistancesensors,etc.

Chapter4FoodForThought

1.Howwouldyoumeasureamotorstorque?Whataboutitsspeed?
YoucanmeasureamotorsspeedbytheRPMsitgenerates.Iamnottoosureabouttorquebutmyguess
isyoucancalculatetorqueifyouknowthePowerandRPMsofamotor.IthinkPower=torquexRPM

2.HowmanyDOFarethereinthehumanhand?
Iamgoingtoassumethewristisnotincludedasthetextbookmentionedthewristwithoutthehand.
Myhandcanoverallmovein3dimensionalspaceandalsoroll,pitch,yawmovementssothats6
DOF.
Eachfingercanmovein3dimensionalspacethemselves,sothats3*5=15DOF
Eachofmyfingershas2joints,sothats10DOF.
Sototal31DOF

Chapter5FoodForThought

HowdoesautomatedWebpathplannerlikeGoogleMapsfindtheoptimalpath?
Well,Irememberfrommycombinatoricsclassthatwecalculatedtheoptimalpathbetween2nodesin
graphusingtheDijikstrasAlgorithm.Iguesssamecouldbeappliedhereasitsprettymuchthesame,the
edgesarestreetsandnodescanbeintersections.

SIKCircuit8
Assemblywaseasy.Ididntscrewintheattachmentfortheservo,asitfitquitesnug.ThistimeImade
sureIgotthewiringrightsoIwouldntscrewituplikethelastcircuit.Forthecode,Ilookedinthe
servo.hlibrarytounderstandthevariousfunctions.Itwasprettybasic.Thecodewasstraightforward
too.Ifirstmadetheservojustgotovariouspositionstoseehowitwouldoperate.Itwaspretty
accurate,thoughIhadtoholditinmyhandthewholetime.ThenIranitwithincrementingpositions.
ThenIranitwithmanycommandssettingitatvariouspositions(Degrees)andtheservowasslow,
seemedliketheprogramwaspushingahead,IguessIneededdelays.
VIDEOhttps://www.youtube.com/watch?v=luoJWE3vTJA

SIKCircuit9
Assemblywasagainstraightforward;Ileftthelastcircuitasitisandaddedtheflexsensor.Avoltage
dividerisusedherejustlikethephotocellcircuit.Andjustlikethatcircuit,thevoltagedivideronlya
returnsaportionofthe0to1023range.Wehavetomapthatrangeto0to180deg.FirstIranthe
programtoseetheresult.Theservoisnotrunningsmoothly,itjitters.Ithencheckedtheoutputinthe
serialwindow.Theservoreadingwasat120,Ichangedthedefaultrangeof600900to726900asthe
flexsensorreadingontheserialwindowstates726.Irantheprogramagainandtheservoreading
startedat0.Nowtoavoidthejitterymovement,Ihadtomakesharp,quickflexingonthesensor.
VIDEOhttps://www.youtube.com/watch?v=urDeeKREW68

SIKCircuit10
Justwantedtonotethattomaketheassemblyprocessmoreeducational,Istudythecircuitguideand
trytoassembleitwithoutreferringtoit.OfcourseIalmostalwaysmakeerrors,butatleastIthinkIam
learningbetter.InthiscaseIputtheGNDwireaccidentallyonthewrongRGBpin,butcaughttheerror
whenIdoublechecked.Ialsohadpushtheresistorsapartastheyweretouchingeachothersincethey
weretooclose.Forthecode,havingdonethepreviouscircuits,Iimmediatelyunderstandthewaytodo
thisone.Wejustreadasensorvaluethroughanaloginputlikebefore.Andjustlikepreviouscircuits,we
havetomapthevoltagerangefromsensor(01023)totheRGBrange(0255).Eachcolor,red,green,
andbluegetsathirdofthetotalvoltagerangeread.Finallywejustoutputthecolors.Rantheprogram
anditworks.Butthesoftpotwastooflimsy,itkeptcomingoutofthebreadboard,andIhadtoholdit
withonehandwhileIpresseditwiththeotherhand.Anotherwaywastopushdownonitwhilesliding
myfingeralongthestrip,thecolorschangegradually.
VIDEOhttps://www.youtube.com/watch?v=vplSec_ODko

SIKCircuit11
Iwasexcitedtobuildthiscircuitandplaywithsound.Thepinsonmybuzzerwereseverelybent.After
workingthembacktoshapeImanagedtosecurelypinittothebreadboard.Therewashardlyanything
elseneededtobuildthiscircuit.Triedtoworkthroughthecodeandunderstanditfirst.Thenotes
correspondingtovariousfrequencieswashelpful!AllIneededtoknowthenwasthelengthofthesong
(numberofnotestoplay),tempo(thespeedtoplayalteredthroughdelays).Twoimportantarrayswere
neededtocustomizemusic,thenotesarraythatactuallyholdsthenotesforthesong,andthebeats
arraythatstateshowlongeachnoteisplayed.Iranthecircuitanditplayedthenotesbeautifully.Ithen
changedthenotestoplayjinglebellsfromwhatIrememberplayinginmusicclass.Iused:Songlength
26,tempo250andthefollowing:
charnotes[]="eeeeeeegcdefffffeeeeddedg";
intbeats[]={1,1,2,1,1,2,1,1,1,1,2,,1,1,1,1,1,1,1,1,1,1,1,1,1,2,2};
Anditplaysbeautifully.
VIDEOhttps://www.youtube.com/watch?v=wA2ObubzPmo

SIKCircuit12
Therewerealotmorethingstoputtogetherforthiscircuitthanthepreviousones.Ithoughtitwould
justbeamatterofgivingpowertothemotor,butapparentlynot.ThenIlearnedthatthepinsonthe
arduinoarenotstrongenoughtorunthemotor,thatmakessense.Ifirstreadandlearnedthevarious
partsofthecircuitbeingused:Forinstance,thediodeisusedtopreventaspikeinvoltagefromthe
motordamagingthetransistor.Idoublecheckedmycircuitcarefullybeforerunningtheprogram.Itran
asexpected.Iattachedapieceofelectricaltapetothemotortoactasamakeshiftpropeller.
VIDEOhttps://www.youtube.com/watch?v=AOPCz9FbT4

MeetingLearningObjectives
1)listthevariousmechanismsofroboticmovement.
Somemechanismsofroboticmovementarearms,Legs,Tracks,Wheels,Tracksandpropellers
2)discussactiveversuspassiveactuation.
Activeactuationusesactivepowerconsumptiontoprovidepower,whereaspassiveactuationuses
potentialenergyintheeffectorsmechanicsanditsinteractionwiththeenvironmenttoprovidepower.
3)describedifferenttypesofactuators.
ThetypesofactuatorsincludeElectricmotors,Hydraulics(usefluidpressure),Pneumatics(useair
pressure),Photoreactivematerials(reacttolight),Chemicallyreactivematerials(likefiberinacid&
alkalinesolutions),Thermallyreactivematerials(reacttotemperature),andPiezoelectricmaterials
(reacttopushingandpressing)
4)discussthesimilaritiesanddifferencesbetweenDCmotors,gearedmotors,andservomotors.
Allofthesemotorsconvertelectricalenergyintomechanicalenergybymeansofusingmagnets,loops,
wire,andcurrenttogeneratemagneticfieldsthatinturnrotatethemotorshaft.Gearedmotorsaddto
aDCmotorbyallowingittochangeforceandtorqueoutput.Thisisparticularlyusefulforrobotsas
robotscantradeaDCmotorsspeedwithmoretorqueinagearedmotor.ServomotorsaresimilartoDC
motorsexceptforthefactthattheycanrotatetheirshafttoaparticularposition.Thisisusefulin
roboticsforprecisecontrolovereffectors.
5)describeandcalculatethedegreesoffreedominmovement.
ADegreeofFreedom(DOF)istheminimumnumberofthecoordinatesneededtodescribeamechanical
systemsmotion.Forinstanceifarobotmovesin3dimensionalspace,themovementinthe3axis(x,
y,z)givesus3DOF.Ifthatrobotcanroll,pitch,oryaw(whichareallrotationsalongthedifferent
axis),thenthats3moreDOF.Themoretheindependentcomponentsthatcanmovelikethis,themore
DOFthatsystemwouldhave.
6)discussstabilityinroboticmovement.
Thereare2typesofstabilityforrobots,staticanddynamic.Arobotthatisstablestaticallycanhold
itselfupwithoutfallingwhileitisnotmovingbyhavingenoughstaticcomponentstosupportitself.A
robotisstabledynamicallyifitneedstomovetobeabletokeepitselfup,likeaflyingrobot.

Unit3
Chapter6FoodForThought
1.HowmanyDOFarethereinthehumanhand?Canyoucontroleachofthemindependently?
Iamgoingtoassumethewristisnotincludedasthetextbookmentionedthewristwithoutthehand.
Myhandcanoverallmovein3dimensionalspaceandalsoroll,pitch,yawmovementssothats6
DOF.
Eachfingercanmovein3dimensionalspacethemselves,sothats3*5=15DOF
Eachofmyfingershas2joints,sothats10DOF.
Sototal31DOF
Yes,youcancontroleachofthemindependently.Although,somemovementsofthepinkiefingercause
minormovementinthenextfinger.
2.Whichofthetwojointtypeswediscussed,rotationalandprismatic,ismorecommonlyfoundin
biologicalbodies?Canyouthinkofexamplesofspecificanimals?
Biologicalbodiestypicallyhavearotationaljoint.Theyhavemusclesaroundajointactingasactuators
toallowthemtofreelymovearoundanaxisonthejoint.Forexampleadagcanmoveitsheadfreely
aroundanaxisparalleltoitslineofsightwhenlookingstraightahead.
3.Areastronautsuitsexoskeletons?Whataboutalevercontrolledbackhoe?
Yes,astronautsuitsareexoskeletons,becausetheywearitanditisanouterstructurethatprovides
protection.Backhoesarenotexoskeletonsbecauseeventhoughapersoncansitinone,theydont
exactlywearit.
4.Imagineanexoskeletonwhichhasitsownsensors,makesitsowndecisions,andactsonthem.Thisis
definitelyarobot,yetitisalsocontrolledbyahumanandattachedtothehumansbody.Canyou
imaginewherethiscouldbeuseful?
Thiscouldbeusefulinhazardousconditions,suchaswarzones,areaofachemicalspill,etc.Thesuitcan
senseexternaldangersandreactaccordinglywithoutthewearersdirection.Itcanforinstance
automaticallysenddatabacktoacentralcomputer,etc.
5.Whataretheendeffectorsofthedogsinthatcase?Arethedogsnowpurelymobilerobots?
Ithinkthattheendeffectorsinthiscasearethedogfronttwolegs,becausethosearethemainlimbs
makingtheimpactonitsgoals.Thebacktwolegsarejustgettingthefronttwolegstotheirtarget.The
dogsarenotpurelymobilerobotsbecausetheirfronttwolegscanbeusedtoperformtasksotherthan
walking.

Unit3Exercises
DesignQuestion
Howmanymotors,andofwhattype,wouldyourequiretomakeafullyfunctionalroboticarmthat
hadaworkingelbow,wrist,andendaffector(i.e.asimpleclamp)?Whatcomponentswouldyouadd
ifyouwantedtheclamptobeabletotellhowharditwasgrabbinganobjectsuchasanegg(i.e.to
avoidcrushingit)?Discussyourdesigninyourweblogindetail,especiallydescribingthechoiceof
motorforeachjoint,thedegreesoffreedomandtherangeofmotion.
Inthefollowingpassage,whenIsawajointrotatesinoneaxis,Imeanthearmmovesalonginoneaxis
likehowafoldedpocketknifeopensup.

Soweneedtosimulateshoulder,Elbow,wrist,clamp
Sofordegreesoffreedom:
1)Clamp:2DOFsincetheneedtobeabletorotateincircularmotionANDgraspthings
2)Wrist:1DOFtorotateinoneaxis
3)Elbow:1DOFtorotateinoneaxislikewrist
4)Shoulder:2DOF,1forrotatinginoneaxislikeelbow,andtheotherDOFforrotatingwholearm
(aroundthebase)
Sototal6DOF,sowewouldneed6Servomotorstoimplementthis.
Theclampjointcanrotate360degrees.
TheWristjointcanrotate180degrees,whichisrealisticandenoughsincetheelbowcanalsomove
allowingthewristtoreachmore
Theelbowcouldbemadetorotateinarangeof180degrees,butitispossibletomaketogoalltheway
aroundupto300degrees(cantdo360sinceitwillcollidewithbaseofarm),itdependsonhowone
wouldwanttoimplementit.Iwoulddo300degreestomakeitmoreversatile

Theshouldercanrotate180degrees,basicallystartingparalleltothefloorononesideandgoingallthe
wayover180degreestobeingparalleltotheflooragainonotherside.
Theshouldercanthenalsorotatealonganaxisperpendiculartothefloor,whichallowsthewholearm
totwist.

Grip:
Wewillalsoneedonemotorfortheclamptogripsomething.Ithinktherewouldbesometypeforforce
feedbacksensorsthatcanbeused.Suchsensorsfromthegripcanrelayinformationbacktothecontrol
unitwhichcanthencontroltheforcebeingapplied.Iamnotsurehowexactlytocontroltheforce,
perhapsthereiswaytovarythetorque.

Programming/CircuitTask
Soobviouslywewillneedaservotoimplementthis.SoIfirstreviewedcircuit8fromtheSIKagainto
recaphowIusedtheservo.Nowtoimplementourelbowmovement,Ijustneedtogettheservoto
movetoaparticulardegreeinputfromuser.(startsat0degree,elbowfullyextended).SoobviouslyI
needaloopthatkeepscheckingforuserinputandmovestheservoaccordingly.GiventheSIKcircuitsI
did,thisisprettystraightforward.Iusedthecircuit8codeasatemplatetoworkfrom.Iknewthepin
assignmentsfortheservo.Iconnectedthepower,groundleadsandtheinputfromtheservotopin9
justasIdidbefore.Everythingworkedrightaway.Ijusthadanifstructuretomakesureuserinputwas
withinlimits.Refertocodecommentsbelowforexplanation.
Testing:Ienteredlimitsoutofrange,andgottheappropriateresponse.IfIentersamedegreevalue
twice,thearmjuststaysput,whichisfine.
CODE:
#include<Servo.h>//servolibrary
Servoservo1;//servocontrolobject
voidsetup()
{
servo1.attach(9);//servoatpin9
Serial.begin(9600);//setbaudrate
servo1.write(0);//moveservoto0degrees
Serial.println("Enteradegreevaluetocontracttheelbowto(0170)");
}
voidloop()
{
intinput;//holdsuserinput
if(Serial.available()){
delay(1000);//Pausetoletservomove(includingwhenitsetsinitial0)
input=Serial.parseInt();//getuserinput
if(input>=0&&input<=170){//checkifdegreewithinlimits
servo1.write(input);//movearm
Serial.print("movingtodegree");//printstatus
Serial.println(input);
}else{//elseifbadinputentered
Serial.println("badinput!");
}
}
}

Connect3jumpercablestoservomotorcabletoserveasextensiontoservomotorcables
GNDwirefromservomotortoe5
5Vcablefromservomotortoe6
Signalcablefromservomotore7
Cablefromc7topin9
Cablefromc6to+
Cablefromc5to
Cablefrom+to5V
CablefromtoGND

VIDEO
http://www.youtube.com/watch?v=iEhIBAsYUCg

Unit4
Chapter7FoodForThought
1.Uncertaintyisnotmuchofaproblemincomputersimulations,whichiswhysimulatedrobotsarenot
veryclosetothereal,physicalones.Canyoufigureoutwhy?
Uncertaintyisnotmuchofaproblemincomputersimulationsbecausesuchsimulationsaretakingplace
inanartificialenvironmentwheremanyvariablescanbecontrolled.Intherealworldtherearemany
uncontrolledvariablesthatincreaseuncertainty.EveninflightsimulationslikeinNASA,theyhaveto
purposelychangesomethingcontrolledtosimulateuncertaintyforthetestpilot.Buteventheycant
possiblyaccountforallunkowns.

2.Argumentthatsensorsarethelimitingfactorinrobotintelligence:
Ofcoursenot,sensorsarenotthelimitingfactor.Nomatterhowmanysensorsyouhave,thecontroller
wouldstillneedtodobetterprocessing.SoIthinkthecontrolleristhelimitingfactorasevenwithfewer
processors,thecontrollerscouldprocessinformationmoreefficiently.Butofcourse,addingmore
sensorsdoesmakeiteasierforthecontrollerastherewillbebetterinformation,butsensorsarenotthe
limitingfactor.
3.Whatdoyouthinkitwilltaketogetrobotstobeselfawareandhighlyintelligent?Andifsomeday
theyareboth,whatwilltheirintelligencebelike,similartooursorcompletelydifferent?
Idontthinkrobotswilleverbeselfawarethewayweare.Ithinkwithenoughtechnological
advancement,theycansimulateselfawarenessorbeselfawaretosomedegree.Samegoesfortheir
intelligence.Themainissueisweourselvesdontfullyunderstandeverythingaboutourconsciousness.
Weknowwhenbabiesareborntheyarenotselfawareorintelligent.Thenafterafewyearsthey
becomeselfawareandtheirintelligencegrowsandtheylearnconstantly.Wedontfullyunderstandthe
underlyingprocessthatmakesallthishappen.Andsincewedontgetit,wecantexpecttobuildrobots
tobethatway.
Chapter8FoodForThought
1.Whymightyoupreferapassivetoanactivesensor?
Imightpreferapassivesensortosaveenergyorpower.Someremoteapplicationsthathavenoaccess
topowerbenefitfrompassivesensorsduetotheirlowenergyusage.
2.Arepotentiometersactiveorpassivesensors?
Potentiometersarepassivesensorsbecausetheydontprovidetheirownsignalorstimuli,theydon't
haveanemitter.Theymeasuretheresistancefromtheuserturningtheknob.
3.Whatrobotsensorswouldyousayaremostsimilartosuchstretchreceptors?Aretheysimilarinform
(mechanismofhowtheydetect)orfunction(whattheydetect)?Whymightstretchreceptorsbeuseful
torobots,evenwithoutstomachsandeating?

Itishardtothinkofthetypeofsensorswithoutknowinghowtheexpandablestomachwillbe
constructed.GiventhetypeofsensorsIhavelearnedthusfar,flexsensorsseemthemostappropriate.
Youcanlinearobotstomachwithablanketofflexsensors,anddetecttheamounttheybend.Such
sensorswouldbeusefulforrobotsbecausetheycanavoidphysicalstressdamageoftherobot.

Chapter9FoodForThought
1.Whatisthespeedofsoundinmetricunits?
Thespeedofsoundinmetricunitsis300m/s.
2.Howmuchgreateristhespeedoflightthanthespeedofsound?Whatdoesthistellyouaboutsensors
thatuseoneortheother?
Thespeedoflightinmetricunitsis299792458m/s.Thereforethespeedoflightisgreaterthanthe
speedofsoundby299792458300=299792158m/s.Thistellsusthatwithlightsensorsyoucanget
informationmuchfasterthansoundsensors.Thereforeprocessingcanbefasterwithlightsensors,
whereaswithsoundsensoryouneedtohavedelaysinbetweenprocessing.
3.Whathappenswhenmultiplerobotsneedtoworktogetherandallhavesonarsensors?Howmight
youdealwiththeirsensorinterference?InChapter20wewilllearnaboutcoordinatingteamsofrobots.
Onewayistotreattheuseofsonarasasharedresource,whereeachrobottransmitsasignalthatitis
pinging.Sotheonerobotexclusivelylockstheuseofsonar.Insteadoftransmittingasignal,Iguess
otherrobotscanlistenforanactiveping,butitcouldbenoisefromtheenvironment.
4.Whydontweuseshiftinfrequenciesbetweensentandreflectedsoundwavesinrobotics?
Iamguessingithastodowiththefactthatitischeaperandsimplertocalculatethetimeofflightas
opposedtoashiftinfrequency.
5.Sincetwoeyesaremuchbetterthanone,arethreeeyesmuchbetter,orevenanybetter,thantwo?
Twoeyesaresignificantlybetterthanoneeyebecauseitgivesasenseofdepth,andthereforelotmore
information.Thirdeyeinthesamedirectionwouldntaddmuchmorebenefit,ifitisveryclosetothe
other2eyes.Ifitisawayfromtheother2eyesinthesamedirection,itcanoffertheadvantageofa
differentangle.Ifthethirdeyeisfacinganotherdirectionthantheother2eyes,thenitcangive
significantlymoreinformationandthusismuchbetter.

Unit5
Chapter10FoodForThought
1.Whathappenswhenyouhavesensorerrorinyoursystem?Whatifyoursensorincorrectlytellsyou
thattherobotisfarfromawallbutinfactitisnot?Whataboutviceversa?Howmightyouaddress
theseissues?
Whenthereisasensorerrorinasystem,therobotwillnotbehaveasexpected.Iftherobotfailstotell
youcorrectlythatitisclosetoawall,itcancollidewithit.Ifitincorrectlytellsyouthatisclosetowall,
thenitcouldstopevenifitissafetomoveforward.Itiscommonforsensorstofailnowandthen,and
weshouldevenaccountforapermanentfailure.Onewaytoaccountforincorrectsensordataistohave
thesensortakemultiplereadingsandaveragetheresults.Thiswilleliminateanyanomalies.Ifthey
sensorrepeatedlygivesincorrectdata,thenwecanhaveabackupsensortopreventacollision.For
instance,ifadistancesensor(IR)ismalfunctioning,aflexsensorinthefrontofarobotcanprevent
collisionbydetectingawall.
2.Whatcanyoudowithopenloopcontrolinyourrobot?Whenmightitbeuseful?
Withopenloopcontrol,arobotexecutescommandsbasedonwhatwaspreviouslypredictedas
opposedtorelyingonsensoryfeedback.Thiscouldbeusefulforrobotswhentheirsensorseitherdonot
orcannotrelyontheirsensors
Chapter11FoodForThought
1.Howimportantistheprogramminglanguage?Coulditmakeorbreakaparticulargadget,device,or
robot?
Anyprogramminglanguagecanbeusedtoimplementsoftwareforarobot.Whatismoreimportantis
howthecontrolarchitectureisusedtoimplementthecontroller.However,certainprogramming
languagescanbebettersuitedforaparticularimplementation.Thiscanmaketheprogrammingeasier,
andthereforethesoftwarepotentiallybetter.Forinstance,theprogramminglanguagePrologcouldbe
wellsuitedforachattingrobot.Thischoiceofprogrammingwillmakethesoftwareimplementation
easierandpotentiallybetter.So,whilethechoiceofprogramminglanguagemightnotnecessarilybreak
arobotordevice,itcouldinhibititspotentialperformance.
2.Withtheconstantdevelopmentofnewtechnologiesthatusecomputation,doyouthinktherewillbe
increasinglymoreorincreasinglyfewerprogramminglanguages?
Ithinkwithdevelopmentofnewtechnologieswewouldseemanynewprogramminglanguages.
However,Ithinkmostofthesenewlanguageswouldbevariantsofcurrentprogramminglanguages,
withminorchanges.Thesenewprogramminglanguageswouldbebeneficialbecausetheycouldbe
tailoredtonewtechnologies,andtherebymakingsoftwaredevelopmentmoreefficient.

Unit6
Chapter12FoodForThought
1.Doyouthinkanimalsuseinternalmodels?Whataboutinsects?
Ithinkanimalsdouseinternalmodels.Whenanimalsareborn,theyhaveunderlyinginstinctstoguide
them.ButIthinkastheyage,theydevelopinternalmodelsbasedontheirexperiences.Iamnotsure
aboutinsects;Ithinktheyarelesscapableofdevelopinginternalmodelsbecauseoftheirbrainsize.
Theystillmayhaveprimitiveinternalmodels.
2.Whymightyounotwanttostoreanduseinternalmodels?
Storingandusinginternalmodelscouldbetaxingintermsofprocessing.Alsoforexample,ifthe
environmentisdynamicorchanging,thebenefitofstoringinternalmodelsactuallybecomesaburden.
Itwouldbebettertonotstoreinternalmodelsandinsteadforexample,userealtimesensorsand
processingtoachievethesamegoal.
Unit7
Chapter13FoodForThought
1.Canyouusedeliberativecontrolwithouthavingsomeinternalrepresentation?
NoIdontthinkyoucandeliberativecontrolwithoutsomeinternalrepresentation.Because,ifarobot
needstobeabletodeliberateorthinkhardaboutwhatitneedstodo,itneedsinformationtoprocess.
Thisinformationhastobestoredinsomemannerwithintherobot,whichwouldbesomesortof
internalrepresentation.However,itispossibleforarobottoreacttorealtimesensoryinputbasedona
setofruleswithoutanunderstandingofitssurroundings.
2.Cananimalsplan?Whichones,andwhatdotheyplan?
Yesmostanimalscanplans,whentheyhavebrainsdevelopedtoacertaindegree.Thisisusually
correlatedwithbrainsize.Forinstance,Lions,KillerWhales,etc.canplanandhuntinapack.
3.Ifyouhadperfectmemory,wouldyoustillneedtoplan?
Ifwehadperfectmemory,Ithinkwewouldstillneedtoplaninsomecasesbecausewemightnothave
experiencedallscenariosinthepasttopurelyrelyonourmemory.Wemighthavetoextrapolate
informationfromourmemorytoplanforacurrentscenariothatisunexpected.
Chapter14FoodForThought
1.Canyouchangethegoalofareactivesystem?Ifso,how?Ifnot,whynot?
No,youcannotchangethegoalofareactivesystembecausetheyhaveaspecificsetofrulesthatcouple
specificscenariostospecificactions.Sincethesefactorsdontchange,thegoalcannotultimately
change.
2.Canyoualwaysavoidusinganyrepresentation/worldmodel?Ifso,how?Ifnot,whynot,andwhat
couldyoudoinstead?
Youcanavoidusingarepresentationmodelbyhavingtherobotrelyonsensoryinput.Gathering
informationinsuchmannerwouldalsoalleviateburdenontherobothavingtostorelargeamountof
data.
3.Canareactiverobotlearnamaze?
Yes,arobotcanlearnamazebystoringaninternalmodelonwhatitencounterswhileexploringallthe
pathwaysofthemaze.Itispossiblefortherobottotraverseallcornersofthemazeiftheyare
accessible,althoughitwouldtakealongtimeandalotofprocessingbytherobot.
Unit8
Chapter15FoodForThought
1.Isthereanalternativetohybridsystems,oristhatallthereisintermsofintelligent,realtimerobot
behavior?Canyouthinkofanotherwaythatarobotcanbeabletoboththinkandreact?
Unfortunately,Ialreadyskimmed(briefly)thebookandknowwhatiscomingahead.However,Istill
knewaboutthismatterhavingwrittensomesimpleAIchatprogramsusingProlog.Otherthanreactive
andhybridsystems,arobotcanhavebehaviorbasedcontrol.Therobotcanbetaughtasetof
behaviorsasopposedtoexplicitdirections.Itcanreactaccordingtothesebehaviorstomeetitsgoals.
Chapter16FoodForThought
1.Somepeoplesaythatbehaviorsshouldbemorepreciselydefinedsothatbehaviorbasedrobot
programmingwouldbeeasiertofigureout.Othersbelievethathavingbehaviorsbelooselydefined
allowsrobotprogrammerstobecreativeandcomeupwithinterestingandnovelideasforwaysto
designrobotcontrollers.Whatdoyouthink?
Ithinkbothstrategiesholdmerit.Topickeitherstrategy,itwouldultimatelydependontheapplication
oftherobot.Iftherobotwasbeingimplementedtomeetsensitiveorcriticalgoals,thenitwouldbe
bettertohavetightlydefinedbehaviorsinordertoavoidunexpectedbehavior.However,iftherobot
hadalesscriticalgoalsuchasanAIchatsystem,thenhavinglooselydefinedbehaviorscouldactually
benefittherobotbecauseunexpectedbehaviorswouldbemoreacceptable.
2.Differencebetweenbehaviorismandbehaviorbasedsystems.
Thedifferencebetweenbehaviorismandbehaviorbasedsystemsisveryprofound.Inbehaviorism,
thereisnoconsiderationgiventoanythinginternaltoasystemorentity.Butbehaviorbasedsystems
canhaveasetofbehaviorsorrulesasfoundationsthatcanshapetheiractionsintheirenvironment.
Sincebehaviorismdiscreditstheseinternaldrivingforces,theyareprofoundlydifferent.

Unit9
Chapter17FoodForThought
1.Doesiteverhappentoyouthatyoucantdecidebetweentwooptions?Whatdoyoudo?
Ihavefacedindecisivenessmanytimes.Sometimesitismorechallengingthanothers.ThefirstthingI
alwaysdointhesesituationsisdeducetheproblemlogically.Theharderitbecomestomakeadecision,
thelessIrelyonlogicandthemoreIrelyongutfeelingorinstinct.Thisinstinctisbasicallyemotion.
Nowrobotscandefinitelybenefitfromsuchinstinctivedecisionmaking.However,thatiseasiersaid
thandone,astheseinstinctsderivefromthecurrentmood,circumstanceorrelationshipwithother
entities.Basically,weareunabletoinstillsuchadvancedmentalcapabilitiesintoaroboticsystem.While
wecanmakearobotmimiccertainthingslikemood,thementalcapabilitiesneededtohaveinstincts
aretoocomplex.So,ifarobotissomedaycapableofsuchinstincts,thenitwoulddefinitelybe
beneficial.

However,suchinstinctscanalsoleadtoimperfections,justlikeinhumandecisionmaking.So,perhaps
wewouldavoidsuchrobotsincertaincriticalapplicationswhereweneedamorerigorouscontrolover
decisionmaking.
Unit10
Chapter18FoodForThought
1.Besideswallfollowing,orfollowing(servoingtosometarget),whichisanaturalbehaviortoproduce
viaemergence,canyouthinkofothersyoumightwanttoimplementthisway?
Anotheremergentbehaviorissearching.Ifyougivearobotasetofrulestoavoidwallsandobjectsina
randommanner(i.e.randomizethedirectionafteravoidingobstacles),thentherobotwouldtechnically
bemappingitsentireenvironmenteventually.Iamnotsurethisisagoodexample,butwithout
explicitlytellingtherobottofollowasetofdirectionstomapitenvironmentorgiveitaninternal
model,randomlyavoidingitsobstaclescouldintheoryeventuallyhavetherobotmapitsentire
environment,thoughinanextremelytediousandinefficientway.

2.Canyouthinkofwaysofusingemergentbehaviortodosomethingquitecomplicated,suchasbuilda
bridgeorawallorabuilding?Antsdoitallthetime,seeChapter20.
Wecanhaveateamofrobotsworkingtogethertoachieveacomplexgoallikebuildingawall.Each
robotcouldbeassignedasimpletasksuchaspassingabrickorlayingabrick.Overall,buildingthewall
wouldthenbeanemergentbehavior.

Unit11
Chapter19FoodForThought
1.Antsaretremendouslyeffectiveatdeadreckoning,findingtheirwayinthevastopenspacesofthe
desertthroughodometry.Theyseemtobemuchbetteratitthanpeople,basedonsomestudiesthat
havebeendone.Ofcourse,nostudieshaveactuallypittedahumanagainstanantinthedesert,but
comparableexperimentsfindpeoplelostandconfused,andantsonanearlystraightpathbackhome.
Antsaresuperiorinnavigationduetotheirinternalcompassandamechanismbywhichtheycanjudge
thedistancetravelled.Itispossibleforhumanstonavigatealmostasgoodasants,althoughnot
naturallyatfirst.Humanscouldusecelestialnavigationtechniquestodeterminetheircurrentlocation
totheirstartinglocation.Theycouldalsokeeptrackofthenumberofstepstheywalkortimetheywalk,
eithermentallyorrecordingitsomewhere.ButIguesstonavigatelikeanant,itwouldhavetobe
mentally.Andofcourse,repeatingthesetechniquesovertimewouldalsoimprovethenavigation.Over
thousandsofyears,itcouldalsobeinherenttohumanbehavior.
2.Agreatdealofresearchhasgoneintounderstandinghowratsnavigate,sincetheyareverygoodatit
(soareants,aswenotedabove,buttheirbrainsaremuchhardertostudy,beingsosmall).Researchers
haveputratsintoregularmazesandwatermazes,andevenmazesfilledwithmilk.Onethingthatis
veryinterestingaboutratsthatlearnamazewellisthattheymayrunintoawall(ifanewwallisstuck
inbytheexperimenter)eveniftheycanseeit,seemingtorunbasedontheirstoredmapratherthan
theirsenseddata.Whattypeofrobotcontrolsystemwouldproducesuchbehavior?Ratsdoitonlyonce,
though.
Arobotcontrolsystemsimilartothiswouldbeahybridsystemthatinvolvesbothdeliberativeand
reactivecontrol.Aratcanclearlydeliberateandhaveaplanorrepresentationofthemazebeforeits
consequentruns.Whenithitsawall,itdeliberatesagaintochangeitsinternalrepresentationorplan
andcontinues.Thus,theratwillnotrunintothatwallagain.

Unit12
Chapter21FoodForThought
1.Somepeoplefindtheideaofrobotslearningfrightening,becausetheyareuncomfortablewiththe
notionofrobotbehaviorbeinganythingbutcompletelypredictable.Doesaddinglearningtotherobots
abilitiesmakeitmoreunpredictable?Ifso,howcanthisbehelped?Ifnot,howcanpeoplesworriesbe
puttorest?
Ifrobottrulyhaslearningasoneofitsabilities,thenitwouldeventuallybehaveinanunpredictable
manner.Unless,ofcourse,itsacontrolledenvironmentwherewecanmonitorwhattherobotlearnsor
controlwhatitlearns.Inadynamicenvironment,learningwouldmakearobotexponentiallycomplex
anddifferentinbehaviorovertime.However,thisneednotbeanissueifwecangivetherobot
fundamentalrulestofollowunderallcircumstances.Thatshoulddampennegativeeffectsofmostof
theunpredictablebehavior.
Unit13
Chapter22FoodForThought
1.Howdoyouthinkthisshouldbejustified?Arerobotstakingtheserolesawayfrompeopleorarethey
fillingotherwiseunfilledroles?Whatifthoserolesareunfillednow,butoncerobotsfillthem,theycould
neveragainbefilledbypeople?Thesearejustsomeoftheethicalandeconomicquestionsthathave
beenposed.Howwouldyouanswerthem?
Ithinkitsamixofeverything.Iamsurerobotsarereplacingcertainhumanlabor,androbotsalsodo
jobsthatareunfilledbyhumanlabor.Ithinkinthefreemarketeconomywehavehere,theultimate
goalshouldtocompletejobxinthemostefficientwaypossible.Ifthatmeansrobotstakingover
humanlabor,thenthatshouldbetheway.Economicstellsusthatthenthehumanworkforcewould
eventuallyadapttosuchshiftsinthemarketbyperhapstrainingmorehumansinskilledjobsthatrobots
cantperform.However,thereisaplaceforethics,andsoIthinkweshouldsoftenthenegativeimpact
onthelaidoffworkforcethroughvariousmethods.Forinstance,wecouldhaveagradual
implementationofrobotsasopposedtoasuddenmassiveoverhaul.Thiswouldgivetheworkforce
moretimetoadapt.

2.Somephilosophershavearguedthatoncerobotsbecomesufficientlyintelligentandselfaware,they
willbesomuchlikepeoplethatitwillbeunethicaltohavethemworkingforus,asitwillcreatearaceof
slaves.Whatdoyouthinkaboutthat?
Thisishardtothinkaboutasitisveryphilosophical.Itdependsonwhatexactlysufficientlyintelligent
means.Ifweassumethatarobotcanbeassmartashuman,havememories,etc.thentheyare
intelligent.Nowwouldthisintelligentrobothavefreewillorasenseofitself?Wouldithavea
conscience?Suchthingsaretoocomplextoevenimaginearobotacquiring,soIwouldassumethatthey
wouldnothavetheseabilities.Inthatcase,sincewewouldbeabletohaveacleardistinctionbetween
humansandtheserobots,itwouldbeOKtonotgivethemequalrights.However,theycouldhaveas
muchrightsasapetwouldhave,whichisinterestingbecausemanypetsarealsointelligent.We
basicallykeeppetslikeslavesinsomemanner.