Autonomous Robotics. CS Fall Amarda Shehu. Department of Computer Science George Mason University

Transcription

1 Autonomous Robotics CS Fall 2016 Amarda Shehu Department of Computer Science George Mason University

2 1 Outline of Today s Class 2 Robotics over the Years 3 Trends in Robotics Research 4 Course Organization Amarda Shehu (485) Outline of Today s Class 2

3 What is a Robot? I, Robot (2004) Amarda Shehu (485) Robotics over the Years 3

4 What is a Robot? I, Robot (2004) Automaton (Greek, autos self + matos thinking, animated, willing ) Amarda Shehu (485) Robotics over the Years 3

5 What is a Robot? from English translation of 1920 play Rossum s Universal Robots by Karel Capek from Czech robotnik (slave), robota (forced labor, drudgery), robotiti (to work, drudge) from Slavic (arabeit) related to German Arbeit (work) Word coined by Capek s brother Josef, who used it initially in a short story Robotics coined in 1941 in a science fiction context by Isaac Asimov I, Robot (2004) Automaton (Greek, autos self + matos thinking, animated, willing ) Amarda Shehu (485) Robotics over the Years 3

6 What is a Robot? from English translation of 1920 play Rossum s Universal Robots by Karel Capek from Czech robotnik (slave), robota (forced labor, drudgery), robotiti (to work, drudge) from Slavic (arabeit) related to German Arbeit (work) Word coined by Capek s brother Josef, who used it initially in a short story Robotics coined in 1941 in a science fiction context by Isaac Asimov 1 A robot may not injure a human being or, through inaction, allow a human being to come to harm. I, Robot (2004) Automaton (Greek, autos self + matos thinking, animated, willing ) Amarda Shehu (485) Robotics over the Years 3

7 What is a Robot? I, Robot (2004) Automaton (Greek, autos self + matos thinking, animated, willing ) from English translation of 1920 play Rossum s Universal Robots by Karel Capek from Czech robotnik (slave), robota (forced labor, drudgery), robotiti (to work, drudge) from Slavic (arabeit) related to German Arbeit (work) Word coined by Capek s brother Josef, who used it initially in a short story Robotics coined in 1941 in a science fiction context by Isaac Asimov 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 conflict with the First Law. Amarda Shehu (485) Robotics over the Years 3

8 What is a Robot? I, Robot (2004) Automaton (Greek, autos self + matos thinking, animated, willing ) from English translation of 1920 play Rossum s Universal Robots by Karel Capek from Czech robotnik (slave), robota (forced labor, drudgery), robotiti (to work, drudge) from Slavic (arabeit) related to German Arbeit (work) Word coined by Capek s brother Josef, who used it initially in a short story Robotics coined in 1941 in a science fiction context by Isaac Asimov 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 conflict with the First Law. 3 A robot must protect its own existence as long as such protection does not conflict with the First or Second Law. Amarda Shehu (485) Robotics over the Years 3

9 What is a Robot? I, Robot (2004) Automaton (Greek, autos self + matos thinking, animated, willing ) Robotics Institute of America: device that automatically performs complicated often repetitive tasks, or a mechanism guided by automatic controls from English translation of 1920 play Rossum s Universal Robots by Karel Capek from Czech robotnik (slave), robota (forced labor, drudgery), robotiti (to work, drudge) from Slavic (arabeit) related to German Arbeit (work) Word coined by Capek s brother Josef, who used it initially in a short story Robotics coined in 1941 in a science fiction context by Isaac Asimov 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 conflict with the First Law. 3 A robot must protect its own existence as long as such protection does not conflict with the First or Second Law. Amarda Shehu (485) Robotics over the Years 3

10 First Robot The Turk / Automaton Chess Player (1770) Turk Amarda Shehu (485) Robotics over the Years 4

11 First Robot The Turk / Automaton Chess Player (1770) Constructed by Wolfgang von Kempelen in 1770 Played many exhibition chess games Solved the knight-tour problem Even played against Benjamin Franklin in France Turk Amarda Shehu (485) Robotics over the Years 4

12 First Fake Robot The Turk / Automaton Chess Player (1770) Constructed by Wolfgang von Kempelen in 1770 Played many exhibition chess games Solved the knight-tour problem Even played against Benjamin Franklin in France Turk... it was a fake, however, human player hid inside machine Amarda Shehu (485) Robotics over the Years 4

13 First Fake Robot The Turk / Automaton Chess Player (1770) Constructed by Wolfgang von Kempelen in 1770 Played many exhibition chess games Solved the knight-tour problem Even played against Benjamin Franklin in France Turk... it was a fake, however, human player hid inside machine Amarda Shehu (485) Robotics over the Years 4

14 First Real Robot Unimate (1961) Created by George Devol Worked on a General Motors assembly line in New Jersey in 1961 Job consisted of transporting die castings from an assembly line and welding these parts on auto bodies Conducted in Robot Hall of Fame in 2003 Amarda Shehu (485) Robotics over the Years 5

15 Trends in Robotics Research: Classical Paradigm Focus on automated reasoning and knowledge representation Perfect world model Closed world assumption: what is not currently known to be true, is false STRIPS (Stanford Research Institute Problem Solver) Amarda Shehu (485) Trends in Robotics Research 6

16 Shakey (Stanford Research Institute, 1966) First mobile robot to reason about its own actions Programs for seeing, reasoning, and acting Triangulating range-finder for sensing obstacles Wireless radio and video camera Used STRIPS to perform block-worlds tasks Conducted in Robot Hall of Fame in the robot Amarda Shehu (485) Trends in Robotics Research 7

17 Trends in Robotics Research: Reactive Paradigm No models: The world is its own, best model Many successes, but also limitations Inspired by biological systems Genghis by Rodney A. Brooks Polly by Ian Horswill Amarda Shehu (485) Trends in Robotics Research 8

18 Trends in Robotics Research: Hybrid Paradigm = Planning + Reactive Combines advantages of previous paradigms World model used for planning Closed loop, reactive control Amarda Shehu (485) Trends in Robotics Research 9

19 Trends in Robotics Research Classical Paradigm (mid 1970s) exact models no sensing necessary Reactive Paradigm (mid 1980s) no models relies heavily on good sensing Hybrid Paradigm (since 1990s) model-based at higher levels reactive at lower levels Probabilistic Paradigm (since mid 1990s) seamless integration of models and sensing inaccurate models, inaccurate sensors Amarda Shehu (485) Trends in Robotics Research 10

20 Robots Today [auto] [bdog] [rhex] [heli1] [heli2] [heli3] [snake] [hand] [asimo] Amarda Shehu (485) Trends in Robotics Research 11

21 Robotics from SyFy to the (Near?) Future Amarda Shehu (485) Trends in Robotics Research 12

22 Robotics from SyFy to the (Near?) Future Which one of these are autonomous? Amarda Shehu (485) Trends in Robotics Research 12

23 Robotics from SyFy to the (Near?) Future Which one of these are autonomous? What is autonomy? Amarda Shehu (485) Trends in Robotics Research 12

24 Robotics from SyFy to the (Near?) Future Which one of these are autonomous? What is autonomy? Does doing things on one s own involve intelligence or rationality? Amarda Shehu (485) Trends in Robotics Research 12

25 Robotics from SyFy to the (Near?) Future Which one of these are autonomous? What is autonomy? Does doing things on one s own involve intelligence or rationality? What does intelligence look like? What about rationality? Amarda Shehu (485) Trends in Robotics Research 12

26 Robotics from SyFy to the (Near?) Future Fooling a human is not enough need proof of consciousness. Amarda Shehu (485) Trends in Robotics Research 13

27 Robotics from SyFy to the (Near?) Future Fooling a human is not enough need proof of consciousness. Figure: Popular culture beyond the Turing Test Ex Machina (2015) Amarda Shehu (485) Trends in Robotics Research 13

28 Robotics from SyFy to the (Near?) Future Fooling a human is not enough need proof of consciousness. Figure: Popular culture beyond the Turing Test Ex Machina (2015) From fooling to manipulating a human as ultimate test. Amarda Shehu (485) Trends in Robotics Research 13

29 This Course Focus on two themes: Motion Planning: How can the robot automatically plan and execute a sequence of motions that avoids collision with obstacles and accomplishes the assigned task? Amarda Shehu (485) Course Organization 14

30 This Course Focus on two themes: Motion Planning: How can the robot automatically plan and execute a sequence of motions that avoids collision with obstacles and accomplishes the assigned task? Localization and Mapping: How can the robot use sensor-based information to determine its own state and model the world? Amarda Shehu (485) Course Organization 14

31 This Course Focus on two themes: Motion Planning: How can the robot automatically plan and execute a sequence of motions that avoids collision with obstacles and accomplishes the assigned task? Localization and Mapping: How can the robot use sensor-based information to determine its own state and model the world? Emphasis on algorithms, analysis, and implementations Amarda Shehu (485) Course Organization 14

32 This Course Focus on two themes: Motion Planning: How can the robot automatically plan and execute a sequence of motions that avoids collision with obstacles and accomplishes the assigned task? Localization and Mapping: How can the robot use sensor-based information to determine its own state and model the world? Emphasis on algorithms, analysis, and implementations Illustrated with practical applications arising in diverse areas such as mobile systems, navigation and exploration, robot manipulation, computer animation, video games, computational biology, and medicine Amarda Shehu (485) Course Organization 14

33 Syllabus Basic Motion-Planning Algorithms and Foundations (3 WKS) Bug Algorithms Configuration Spaces Forward and Inverse Kinematics for Manipulators Potential Fields Roadmaps/Cell Decompositions Sampling-based and Probabilistic Motion Planning (3WKS) Roadmap Approaches Tree Approaches Advanced Motion Planning (4 WKS) Multiple Robots Manipulation Planning Dynamics/Physics Game Engines Dynamic Environments/Uncertainty Localization and Mapping (2 WKS) Kalman Filtering/Bayesian Methods Mapping and SLAM Amarda Shehu (485) Course Organization 15

34 Textbook Required Recommended (available online, free) Amarda Shehu (485) Course Organization 16

35 Grading Homeworks(3) (45%) Exams(2) (30%) Final Project (25%) Contact Information Class: Art and Design Bldg L008 M/W 12:00-1:15 pm Office: Engineering Building Office Hours: M/W 11:00-11:59am Amarda Shehu (485) Course Organization 17