5. Situated Agents (Robots) Part 1: Introduction to Robotics. ) - PDF document

5. Situated Agents (Robots) Part 1: Introduction to Robotics. ) Vision and uncertainty Vision and uncertainty. tems (SMA-UPC Javier Vázquez-Salceda SMA-UPC Multiagent Syst 16/07/2012 https://kemlg.upc.edu Mobile Robotics  “Robotics is an application area of AI where theoretical solutions have to cope with real problems”  Problems in perception (incomplete, uncertain, noisy) p p ( p , , y) ents (Robots)  Problems in motion (drift, slippage, motion dynamics, obstacles)  In Mobile Robotics some of those problems increase  Large-scale space (regions of space larger than those observed from a single vantage point  Local sensors  Local sensors 5. Situated Ag  Need for • space representation • positional error estimation • Object and place recognition • Real-time response jvazquez@lsi.upc.edu 2

) tems (SMA-UPC Introduction to Robotics terminology Multiagent Syst https://kemlg.upc.edu The Syllabus  Sensors  Inertia  Vision  Torque  Actuators  Actuators  Compass  Compass ents (Robots)  Motion  Joint  (Forward/Inverse)  Bumpers Kinematics  Landmark  Drift  Geometric Map  Localization  Topological Map  Navigation 5. Situated Ag  Basic behaviours  Complex behaviours  Multi-robot behaviours jvazquez@lsi.upc.edu 4

Types of robots (I)  Static Robots vs Mobile Robots ents (Robots) Lunokhod 5. Situated Ag jvazquez@lsi.upc.edu 5 Types of robots (II)  Wheeled Robots VS Legged Robots Sojourner (NASA) ents (Robots) 5. Situated Ag jvazquez@lsi.upc.edu 6

Types of robots (III) • Microbots,  Robots, ents (Robots) – Small, cheap robots, – cheap sensors (no sonar or laser) • Nanobots N b t 5. Situated Ag jvazquez@lsi.upc.edu 7 Types of robots (IV)  Operational regimes  Completely autonomous Spirit (Mars)  Semi-autonomous • Telerobotic ents (Robots) • Teleoperated Lunokhod (Moon) 5. Situated Ag jvazquez@lsi.upc.edu 8 16/07/2012

Back to theory: Wumpus World (I) Actions: 1 breeze hole breeze A • Forward  , • 90 0 right , ents (Robots) 2 smell breeze • 90 0 left, 3 breeze W hole • Shot, • Pick-up, breeze 4 smell • Leave, breeze breeze hole breeze 5 smell • Out Out 5. Situated Ag gold 1 2 3 4 5 Hypothesis 1: discretized world. Hypothesis 2: totally deterministic actions. jvazquez@lsi.upc.edu 9 Back to theory: Wumpus World (II)  Agent cannot perceive anything in its own position  In the square where the wumpus is and the 4 adjacent ones (non diagonal) the agent will perceive an smell (s=1) (non diagonal) the agent will perceive an smell (s=1). ents (Robots)  In squares adjacent to a hole, the agent will perceive a breeze (b=1)  The square containing he gold will show a glitter (g=1)  When the agent smashes against a wall, it will perceive a bump (u=1) bump (u=1) 5. Situated Ag  Perceptions are expressed in lists [ smell (s), breeze (b), glitter (g), bump(u), cry(c) ] Hypothesis 3: limited perception but perfect, without noise. jvazquez@lsi.upc.edu 10

Back to theory: Wumpus World (III) ok v ok v 1 [s,nil,nil,nil,nil] A hole? A A A b ok v wumpus? ents (Robots) 2 A = agent A hole? s ok= safe wumpus? 3 v = visited s = smell 4 b = breeze 5 g = glitter g g 5. Situated Ag u = bump 1 2 3 4 5 c = cry What now? jvazquez@lsi.upc.edu 11 Back to theory: Wumpus World (III) ok v ok v 1 hole? [s,nil,nil,nil,nil] b ok wumpus? ents (Robots) 2 A = agent A hole? s ok= safe wumpus? 3 v = visited s = smell 4 b = breeze 5 g = glitter g g 5. Situated Ag u = bump 1 2 3 4 5 c = cry Memory: In [2,1] there was no s => no wumpus in [2,2] jvazquez@lsi.upc.edu 12

Back to theory: Wumpus World (III) ok v ok v 1 [s,nil,nil,nil,nil] hole? b ok ents (Robots) 2 A = agent A hole? s ok= safe wumpus? 3 v = visited s = smell 4 b = breeze 5 g = glitter g g 5. Situated Ag u = bump 1 2 3 4 5 c = cry In [1,2] there was no b => no hole in [2,2] jvazquez@lsi.upc.edu 13 Back to theory: Wumpus World (III) ok v ok v 1 hole? [s,nil,nil,nil,nil] b ok A ok ents (Robots) 2 A = agent s ok= safe 3 Wumpus! Wumpus! v = visited s = smell 4 b = breeze 5 g = glitter g g 5. Situated Ag u = bump 1 2 3 4 5 no wumpus in [1,1] ^ c = cry no wumpus in [2,2] ^ wall in [0,2] ^ smell in [1,2] ^ I have heard no cry => Wumpus alive in [1,3]!!! jvazquez@lsi.upc.edu 14

Back to theory: Situation Calculus  Perceptual uncertainty problem  solution: Situation Calculus  Allows the description of the world as a sequence of Allows the description of the world as a sequence of ents (Robots) situations , and each one as a snapshot from a world state.  Problems:  Based in hypothesis 1 (discretizable environment)  Additional hypothesis: environment won’t change without Additional hypothesis: environment won t change without 5. Situated Ag my action •  clock is driven by my actions. Hypothesis 4: static environment (it won’t change if I do not change it) jvazquez@lsi.upc.edu 15 Back to theory: Localization Heading(Agent, S 0 ) = 0 o next-location  p,l, S At(p,l, S )  next-location(p, S ) = ents (Robots) direction(l, Heading (p, S )) Adjacency  l 1 ,l 2 adjacent(l 1 ,l 2 )  d  l 1 =direction(l 2 ,d) Wall   x,y wall([x,y])  (x=0  x=lim  y=0  y=lim) 5. Situated Ag Problem: based in perfect knowledge about initial position + hypothesis in totally determinisic actions  Hypothesis 5: perfect knowledge about actual location. jvazquez@lsi.upc.edu 16

Back to theory: Environment properties Real world: Robots Theory: Wumpus World  Partially accesible  Partially accesible Partially observable y Partially observable y   environment, but perfect environment and inperfect ents (Robots) perception perception (noise)  Deterministic  Stocastic Predictable effect for Unpredictable effect for   actions actions (inertia, drift, slipage)  Sequential (non episodic)  Sequential (non episodic) 5. Situated Ag Cumulative errors   Static  Dynamic Clock driven by my  actions Real time   Discrete  Continuous jvazquez@lsi.upc.edu 17 ) tems (SMA-UPC Robot architectures Multiagent Syst https://kemlg.upc.edu

Levels of abstraction COMPUTATIONAL LEVEL COMPUTATIONAL LEVEL Perception Cognition Action ents (Robots) DEVICE LEVEL DEVICE LEVEL Sensor Drivers Actuator Drivers Or Com m unication Or I nterface Sensing Libraries Motion Libraries PHYSICAL/HARDWARE LEVEL PHYSICAL/HARDWARE LEVEL 5. Situated Ag Sensors Com m HW Actuators External World jvazquez@lsi.upc.edu 19 Intelligent Robot (I) Tasks Perception Cognition Action ents (Robots) Sensors Com m HW Actuators External World External World 5. Situated Ag jvazquez@lsi.upc.edu 20

Intelligent Robot (II) Tasks  Perception  sensing, modeling of the world i d li f th ld ents (Robots)  Communication ( listening )  Cognition  behaviors, action selection, planning, learning  multi-robot coordination, teamwork  response to opponent, multi-agent learning  Action 5. Situated Ag  motion, navigation, obstacle avoidance  Communication ( telling ) jvazquez@lsi.upc.edu 21 Intelligent Robot (III) Architectural Paradigms ents (Robots) 5. Situated Ag jvazquez@lsi.upc.edu 22

Intelligent Robot (III) Hierarchical Paradigm  Used in early times of robotics  Problem: time of reaction  Example: Shakey (Standford Univ, 1970) ents (Robots) 5. Situated Ag jvazquez@lsi.upc.edu 23 Intelligent Robot (III) Reactive Paradigm  Reactive paradigm organizes the components vertically so that there is a more direct route from sensors to actuators.  Schematically Brooks (1986) depicts the paradigm as y ( ) p p g follows: ents (Robots) 5. Situated Ag  Problem: conflicting orders to Actuators jvazquez@lsi.upc.edu 24

Intelligent Robot (III) Brooks’ Subsumption Architecture  Components behaviors are divided into layers (modules) with inputs, outputs and a reset.  Arbitration scheme: a module at a higher level can g  suppress the input of a module at a lower level thereby ents (Robots) preventing the module from seeing a value at its input.  inhibit the output of a module at a lower level thereby preventing that output from being propagated to other modules. 5. Situated Ag » Problem: complex set-up of modules to 16/07/2012 jvazquez@lsi.upc.edu 25 avoid low-level reaction problems ents (Robots) 5. Situated Ag jvazquez@lsi.upc.edu 26

Intelligent Robot (III) Hybrid Architectures  Tries to equilibrate deliberation and reactivity  Usually deliberation UNLESS immediate reaction is needed ents (Robots) CONTROL LAYER CONTROL LAYER Plan REACTIVE LAYER REACTIVE LAYER Sense Sense Act Act 5. Situated Ag PHYSICAL LAYER PHYSICAL LAYER Actuators Sensors jvazquez@lsi.upc.edu 27 Intelligent Robot (III) Layers SOCIAL LAYER SOCIAL LAYER INTELLIGENCE LAYER INTELLIGENCE LAYER INTELLIGENCE LAYER INTELLIGENCE LAYER ents (Robots) CONTROL LAYER CONTROL LAYER REACTIVE LAYER REACTIVE LAYER 5. Situated Ag PHYSICAL LAYER PHYSICAL LAYER jvazquez@lsi.upc.edu 28