Transcription
Introduction to RoboticsCS 445 (lab course, 4 units)Amin AtrashLecture #1: Defining RoboticsIntroduction to Robotics L. Itti, M. J. Mataric’1
Lecture OutlineDefining “robot”What makes a robotSensors, sensor spaceState, state spaceAction/behavior, effectors, action spaceThe spectrum of controlIntroduction to Robotics L. Itti, M. J. Mataric’2
Why “robot”?The term “robot” comes from Karel Capek’s1921 play RUR (Rossum’s Universal Robots).It is most likely a combination of “rabota”(obligatory work) and “robotnik” (serf).The kind of robotics we will talk about willmove far beyond such “obligatory work.”Introduction to Robotics L. Itti, M. J. Mataric’3
Who said “robotics”?The term “robotics” comes formCapek’s “robot”, and was firstintroduced by Isaac Asimov in hisscience fiction writing.Introduction to Robotics L. Itti, M. J. Mataric’4
Alternative termsUAV: Unmanned Aerial VehicleUGV: Unmanned Ground VehicleUUV: Unmanned Undersea(underwater) VehicleAUV: Autonomous UnderwaterVehicleIntroduction to Robotics L. Itti, M. J. Mataric’5
What is a Robot?A robot is a system which exists in thephysical world and autonomouslysenses its environment and acts in it toachieve some goals.Introduction to Robotics L. Itti, M. J. Mataric’6
Other DefinitionsA robot is a re-programmable, multifunctional, manipulator designed tomove material, parts, or specializeddevices though variable programmedmotions for the performance of a task(Robotics Industry Association)Robotics is the intelligent connection ofperception to action (M. Brady)Introduction to Robotics L. Itti, M. J. Mataric’7
Anthropomorphic Robots(Having human form or attributes)Introduction to Robotics L. Itti, M. J. Mataric’8
Animal-like RobotsIntroduction to Robotics L. Itti, M. J. Mataric’9
RHEX (U. Michigan)Introduction to Robotics L. Itti, M. J. Mataric’10
Animal-robotsIdea: use biological “mechanics”But endow them with alternative robotic control Cutting edge but poses ethical issues!Applications: search and rescue in rumble,surveillance, etc.Introduction to Robotics L. Itti, M. J. Mataric’11
Talwar et al., 2002(State U. NY)Introduction to Robotics L. Itti, M. J. Mataric’12
RoboratNote: is it really a robot?The rat is roaming freely but its autonomy is limitedbecause cues and rewards are delivered by a humanoperator. They could be delivered by an A.I. system, though Introduction to Robotics L. Itti, M. J. Mataric’13
Roborat lives!Introduction to Robotics L. Itti, M. J. Mataric’14
Unmanned VehiclesIntroduction to Robotics L. Itti, M. J. Mataric’15
What Makes a Robot?A robot consists llerA robot is capable of:acting autonomouslyachieving goalsIntroduction to Robotics L. Itti, M. J. Mataric’16
Rational Agent / link to sIntroduction to Robotics L. Itti, M. J. Mataric’17
Sensors: what can be sensed?Depends on the sensors on the robotThe robot exists in its sensor space: allpossible values of sensory readingsAlso called perceptual spaceRobot sensors are very different frombiological onesA roboticist has to try to imagine the worldin the robot’s sensor spaceIntroduction to Robotics L. Itti, M. J. Mataric’18
State: what can be known?A sufficient description of the systemCan be:Observable: robot always knows its stateHidden/inaccessible/unobservable: robotnever knows its statePartially observable: the robot knows apart of its stateDiscrete (e.g., up, down, blue, red)Continuous (e.g., 3.765 mph)Introduction to Robotics L. Itti, M. J. Mataric’19
Types of StateExternal state: state of the worldSensed using the robot’s sensorsE.g.: night, day, at-home, sleeping, sunnyInternal state: state of the robotSensed using internal sensorsStored/rememberedE.g.: velocity, moodThe robot’s state is a combination of itsexternal and internal state.Introduction to Robotics L. Itti, M. J. Mataric’20
State and IntelligenceState space: all possible states the systemcan be inA challenge: sensors do not provide state!How intelligent a robot appears is stronglydependent on how much it can sense aboutits environment and about itself.Introduction to Robotics L. Itti, M. J. Mataric’21
Internal ModelsInternal state can be used to remember informationabout the world (e.g., remember paths to the goal,remember maps, remember friends v. enemies, etc.)This is called a representation or an internal model.Representations/modelshave a lot to do with howcomplex a controller is!Introduction to Robotics L. Itti, M. J. Mataric’22
Action/ActuationA robot acts through its actuators (e.g., motors),which typically drive effectors (e.g., wheels)Robotic actuators are very different frombiological ones, both are used for:locomotion (moving around, going places)manipulation (handling objects)This divides robotics into two areasmobile roboticsmanipulator roboticsIntroduction to Robotics L. Itti, M. J. Mataric’23
Actuators and DOFMobile robots move around using wheels, tracks, or legsMobile robots typically move in 2D (but note thatswimming and flying is 3D)Manipulators are various robot armsThey can move from 1 to many DThink of the dimensions as therobot’s degrees of freedom (DOF)Introduction to Robotics L. Itti, M. J. Mataric’24
Action v. BehaviorBehavior is what an external observer sees a robot doing.Robots are programmed to display desired behavior.Behavior is a result of a sequence of robot actions.Observing behavior may not tell us much about theinternal control of a robot. Control can be a black box.Introduction to Robotics L. Itti, M. J. Mataric’25
AutonomyAutonomy is the ability to make one’s owndecisions and act on them.For robots, autonomy means the ability to senseand act on a given situation appropriately.Autonomy can be:complete (e.g., R2D2)partial (e.g., tele-operated robots)Introduction to Robotics L. Itti, M. J. Mataric’26
ControlRobot control refers to the way in whichthe sensing and action of a robot arecoordinated.The many different ways in which robotscan be controlled all fall along a welldefined spectrum of control.Introduction to Robotics L. Itti, M. J. Mataric’27
Spectrum of ControlIntroduction to Robotics L. Itti, M. J. Mataric’28
Control ApproachesReactive ControlDon’t think, (re)act.Behavior-Based ControlThink the way you act.Deliberative ControlThink hard, act later.Hybrid ControlThink and act independently, in parallel.Introduction to Robotics L. Itti, M. J. Mataric’29
Food for ThoughtAre exo-skeletons robots?Is HAL a robot?Some intelligent Web agents are called“softbots”. Are they robots?Most, if not all, of the robots you build inthis class will use reactive control. Whatmore is there?Introduction to Robotics L. Itti, M. J. Mataric’30
The term “robot” comes from Karel Capek’s 1921 play RUR ( Rossum’s Universal Robots). . Capek’s “robot”, and was first introduced by Isaac Asimov in his science fiction w
