Construction of Mobile Robots

Transcription

1 Construction of Mobile Robots Institute for Software Technology 1

2 Previous Years Conference Robot Breakfast Robot 2

3 This year.. again 3

4 Industry 4.0 Flexible Production yaskawa wikipedia today robots are deployed in very static setups product customization hard to achieve future robots are deployed in very dynamic setups products and production lines may change quickly product customization easy to achieve also direct cooperation with humans more flexibility needed customization, e-commerce tighter connection of design, production, supplier, hot TU Graz Smart Production Graz 4

5 Interesting Research Challenges epfl across all disciplines mechanical engineering, electrical engineering, computer science, robotics, vision, AI, economy, in particular for us planning and scheduling high-level control and executive perception monitoring and diagnosis mobile manipulation 5

6 Industry 4.0 Research and Fun RoboCup is a competition to promote robotics and AI besides soccer and rescue it deals also with I4.0 RoboCup Logistics League team of robots produces a series of individual products Festo mockup machines (MPS) 6

7 RoboCup Logistics League - Overview Referee Box Team Server 7

8 Setup Game Phases robots and machines will be prepared for the game Exploration robots explore the game area for machines and their type machines are reported to the referee box Production referee box announces product orders incrementally robots derive a production schedule robots setup machines for the proper refinement step robots provide and retrieve intermediate/final products to/from machines 8

9 Robots team of 3 robots Festo Robotino 3 (version 2 phase out) omni-directional onboard linux-based computer additional onboard computer possible additional digital/analog IOs festo 9

10 Our Robots Festo Robotino V2 WiFi access point Ethernet for additional computer bumper, IR distance, propritary electric gripper (products) webcams for tag and light recognition laser scanner for navigation 10

11 Machines Modular Production System (MPS) by Festo Didactics same conveyor interface for in/out can be identified by AR tags controlled by an PLC get commands from referee box different types 1. base station dispenser 2. cap station mounts a cap 3. ring station mounts a ring of a color 4. delivery station final place 11

12 Products products represented by stacks of bases, rings and caps product complexity is determined by number of items and the color adding items resembles a refinement step eventually adding of additional material is necessary 12

13 GRIPS Graz Roust Intelligent Production System TUG student team for RoboCup Logistics League emerged from 2015 course 3 rd place at world championship 2016 in the first year 13

14 Two Tasks two different task for the 2016 assignment students join one team working on one task 1. optimize the exploration (real robots) based on the existing exploration software improve the exploration strategy to speed up exploration distribution of exploration subtasks to robots immediate integration of valuable percepts 2. cooperative production (simulation) based on the existing task assignment software autonomously produce different products improve the central task decomposition and scheduling add or modify production behaviors (subtasks) 14

15 Simplified Environment 6 simple mockups for machines boxes with AR tags and signal lights full production scenario in simulation further simplifications possible if needed 15

16 Audience/Prerequisites Master Information and Computer Engineering/Computer Science/Software Development and Business Management compulsory for major mobile robots interested students specialization for mobile robots expected skills teamwork independent work experience with linux experience with ROS programming skills (e.g. C++) courses: mobile robots/advanced robotics 16

17 Ressourcen robots will be adapted and remain the same for the rest of the course groups can reserve the robots exclusively for slots of a half day, principle first come - first serve, the actual schedule can be found at the google-calendar of the robots powerful simulation server available in the lab access to the lab during late hours or holidays with the TUG Card 17

18 Available Software/Hardware Robots Festo Robotino V2 Software ROS Navigation Stack Point Cloud Library (PCL) AR Tag Library OpenPRS various 3 rd Party Libraries Additional Sensors cameras Kinect Additional Equipment & Small Parts can be purchased or built during the course 18

19 Goal/Procedure students are able to select, build and apply a mobile robot for a task experience with mobile robots, actuators and sensors solve a task with a mobile robot working in small groups (~ 3-5 students) as much as possible freedom und creativity, no usual static lab setup independent working in the lab FUN 19

20 Roadmap , 10:00-12:00, seminar room IST: introduction , 10:00-12:00, seminar room IST : software , 09:00-12:00, robot lab: practical introduction robot, ROS, navigation, OpenPRS, grips architecture , 14:00-17:00, robot lab: practical introduction, Gazebo, RCLL simulation , 23:59: submission design document :00-12:00, robot lab: discussion design document , 10:00-12:00, robot lab: 1. midterm meeting , 10:00-12:00, robot lab: 2. midterm meeting , 10:00-12:00, robot lab: presentation and demo 20

21 Design Document one per task short description of the planned solution content solution idea software modules including 3 rd party libraries robot adaptions including sensors and small parts interfaces to other hardware/software modules brief project plan (in keywords, some milestones) name und id of group members document PDF - 2 pages via to steinbauer@ist.tugraz.at, subject KMR 2016 Design 21

22 Developed Software have to be based on ROS Indigo and the existing TUG RCLL software architecture all freely available ROS packages can be used needed Packages of TUG will be available in the robot lab git repository developed software has to be organized as independent ROS catkin packages including launch and parameter files the software has to be stored in the robot lab git repository access data to the git repository will be provided during the practical introduction 22

23 Grading Task Content Available Points Design Document Solution Proposal Midterm Meeting Presentation of the Project Progress 2. Midterm Meeting Presentation of the Project Progress Final Demo Presentation of the developed solution Live Demo Sum

24 Grading Points Grade -100 % 1-87,5 % 2-75 % ,5 % 4 < 50 % 5 24

25 Information Webpage of the Course: Literature: Papers, Documentation, recommended Books Contact: personal: Inst. for Software Technology, Inffeldgasse 16b/II personal: Robot Lab, Inffeldgasse 13/V Newsgroup: tu-graz.kmr 25

26 RoboCup 2017 we plan to send a student team to participate in the world cup interested students can continue to work on a competitive team integration in the studies projects master theses improvement and completion of hardware/software solution during summer term 26

27 Questions? 27