Intelligent Robotics Assignments

Transcription

1 Intelligent Robotics Assignments Luís Paulo Reis Assignment#1 Oral Presentation about an Intelligent Robotic New Trend Groups: 1 to 3 students 8 15 Minutes Oral Presentation Slides (including appropriate images and videos) 4 Presentationsineachofthenext3 or4 lessons Deadline: October 23th (Extended: November 6th) 2 1

2 Assignment#1 - Subjects Tema 1: Robôs Humanóides: Asimo, Cog, QRIO Tema 2: O Kit NXT (Lego MindStorms) Tema 3: Kits Robóticos Tema 4: Lojas On-line de Material Robótico e Plataformas Robóticas: Com rodas e pernas Tema 5: Mascotes Robóticas ( Robotic Pets ): Tamagotchi, Furby, Techno, Poo-Chi, Furby2 e Outros Tema 6: Automóveis Robóticos Inteligentes, Condução Autónoma e o DARPA Grand Challenge Tema 7: RoboOlympics, Manitoba Robot Games e Eventos Semelhantes Tema 8: Robots de Combate (Battlebots RobotWars e Outras Competições Robóticas Semelhantes) Tema 9: Competições de Futebol Robótico FIRA 3 Assignment#1 Subjects Tema 10: Manipuladores Robóticos Tema 11: Visões de Filmes e Livros sobre a Robótica no Futuro Tema 12: Simuladores de Robótica Móvel Tema 13: UAVs - Unmanned Aircraft Vehicle Tema 14: Robótica Submarina Tema 15: Cadeiras de Rodas Robóticas Tema 16: Robôs reconfiguráveis Tema 17: Robôs na Exploração de Marte Tema 18: Swarming Robotics Tema 19: MicroRobótica e NanoRobótica Tema 20: Microsoft Robotics Studio Tema 21: Novos sensores em Robótica Tema 22: Domótica vs Robótica 4 2

3 Assignment #2 Simple Reactive Architecture General Description Installandtesttheciber-mousesimulator( Test a simple source code(c++ avalaible with simulator source, Java, etc) Develop a simple Subsumption Architecture for a Ciber-Mouse Agent: if at the targetsignal end (1) if detect an obstacle then change direction (2) if not at the target then rotate to find the target (3) if true then moves in front/randomly (4) Change the architecture using more elaborated behaviors (for example: follow a given wall). Write a Paper (3/4 pages, Springer LNCS format) showing which mazes may be solved (and which can t) by this simple architecture. Deadline: October 23th (Extended: October 27th) 5 Assignment #2 Simple Reactive Architecture Objectives: Familiarize with the Ciber-Rato Robot Simulation Tools Use the subsumptionagent architecture for the development of a maze solver agent Acknowledge the benefits and limitations of this architecture (which mazes can be solved and which can t) Description of the assignment: 1. Download the latest Ciber-Rato Source Code Tools from 2. Install the Ciber-Rato Tools in your PC 3. Run Simulator, Viewer and one of the Sample Robots using the default maze 4. Run the simulator from the command line using option -log <filename> and execute a test using three virtual robots 5. Use the Logplayer (and Viewer) to play the logfile that has been saved in the previous step. 6. Develop an agent, using the subsumption architecture, that is able to accomplish the first goal (find the beacon and get there) of the Ciber-Rato Simulation environment. 7. Write a report on the approach that has been developed to tackle the previous step. 3

4 Assignment #2 Simple Reactive Architecture Deliverables: 1. Source code of the developed agent 2. Report (Paper 3/4 pages, Springer LNCS style) describing your implementation of the subsumption architecture for this particular problem Recommended Readings: Ciber-Rato2008 -RulesandTechnicalSpecifications, DETI -Universidade de Aveiro, 2008, "A Robust Layered Control System for a Mobile Robot", Rodney A. Brooks, IEEE Transactions on Robotics and Automation, 2(1), pages 14-23, April Part I Robotic Paradigms of An Introduction to AI Robotics, Robin R. Murphy, Bradford Book, MIT Press, Cambridge, Massachussets, London England, ISBN: Behavior-Based Robotics, Ronald C. Arkin, MIT Press, 1998, ISBN Assignment #3 Development of a Path Planning and Plan Execution Agent Objectives: Develop a Path Planning and Plan Execution procedures for a virtual robot assuming a known maze. Develop a Mapping procedure to acquire an internal representation of a maze as the robot walks around. Integrate the Mapping, Path Planning and Plan execution procedures to develop a deliberative agent that can find its way from the starting position to the target position without prior knowledge of the maze. Deadline: November 20th 8 4

5 Assignment #3 Development of a Path Planning and Plan Execution Agent Description of the assignment: 1. All steps will use the Ciber-Rato Simulation Environment. 2. Assuming your robot knows the maze and its starting position (it can read the files that describe the maze and the starting grid) use A*, or another path planning algorithm to find a path that takes the robot to the target. 3. Develop an agent that is capable of executing the pre-determined path in the noisy and dynamic environment of the Ciber-Rato Simulator. 4. Develop a mapping procedure that allows the robot to acquire an internal representation of the maze as it traverses it. 5. Integrate the path planner and the mapping procedures in order to have an agent that is able to find its way to the target in an initially unknown maze. 9 Assignment #3 Development of a Path Planning and Plan Execution Agent Deliverables: 1. Source code of the developed agent 2. Paper (Springer LNCS Format 4/6 pages) describing your implementation of the path planning, path execution and mapping procedures and of the final integrated agent. Recommended Reading: Chapters 9, 10 and 11 of An Introduction to AI Robotics, Robin R. Murphy, Bradford Book, MIT Press, Cambridge, Massachussets, London England, ISBN: Part II Localization and Part III Mapping of Probabilistic Robotics, Sebastian Thrun, Wolfram Burgard, Dieter Fox, MIT Press, Cambridge, Massachussets, London England, ISBN: Map-based navigation in mobile robots: II. A review of map-learning and path-planning strategies, Jean- Arcady Meyer, David Filliat, Cognitive Systems Research, Volume 4, Issue 4, December 2003, Pages Lists of CiberMouse competition publications at:

6 Assignment #4 Agent Team for Participation in ( Paper (14 Pages Springer LNCS Based on previous papers) Oral Presentation (10-15 min) Or Other Robotic Assignment (discuss with the teacher ) Paper (10/14 Pages Springer LNCS) Oral Presentation (10-15 min) Or Final Examination (3 Hours Practical + Theoretical - All Topics) Or Only for PRODEI: Other Intelligent Robotics -Scientific work (Discuss with the teacher ) Deadline: December 11th (Draft), December 16th (Final), December 18th (Presentation) 11 6