Humanoid Robotics (TIF 160)

Transcription

1 Humanoid Robotics (TIF 160) Lecture 1, Introduction and motivation to humanoid robotics

2 What will you learn? (Aims) Basic facts about humanoid robots Kinematics (and dynamics) of humanoid robots Motion of humanoid robots Locomotion Other motions Human-robot interaction Text-to-speech synthesis (Speech recognition) Image processing (e.g. face tracking, hand-eye coordination)

3 What will you learn? (Aims) Brain structures for humanoid robots Robotic behaviors Decision-making Humanoid robot programming Microcontroller programming (Pbasic) High-level programming (C#) Putting it all together: programming and steering humanoid robots

4 Course work Theoretical exercises (to a rather small extent) Practical work with humanoid robots (complete or partial)

5 Why should you take this course? Humanoid robots are becoming increasingly important, and will most likely have a great impact on society in the foreseeable future (10-20 years). Humanoid robotics combine several different disciplines: Mechanics, electronics, programming, human-robot interaction, ethology etc. Thus, by studying humanoid robots, you learn a great deal about system engineering. Humanoid robots are interesting and fun to work with.

6 Course contents v.1-2 Lectures Theoretical assignments Learning C# programming Learning PBasic programming v.3-7 Robot work (in the ET-lab) Assignments

7 Course contents (v.1-2) Date Time Room Contents FL52 Introduction and motivation FL74 Kinematics, humanoid robot motion FL74 HRI* (1) Speech synthesis and recognition FL52 HRI (2) Image processing and object rec FL74 Programming humanoid robots (1) FL74 Programming humanoid robots (2) * HRI = Human-robot interaction

8 Kinematics and dynamics Direct (forward) kinematics: Determining the position and orientation of a robot part (e.g. a hand), given the joint angles. Inverse kinematics: determining the joint angles, given the robot s posture. Dynamics: The response of the robot to the forces acting on it.

9 Humanoid robot locomotion Kondo KHR-1

10 Humanoid robot motion Hubert

11 Humanoid robot motion Kondo KHR-1

12 Human-robot interaction (HRI) The aim of HRI research is to generate a robot capable of interacting with people in a natural way. HRI involves topics such as Scene interpretation (object detection, face detection) Face recognition Emotion recognition Face detection, object detect Speech synthesis and recognition Dialogue: Analysing input (auditory, visual), making sense of the information (cognitive processing), and responding in an appropriate way

13 Robot brain structure One of the most challenging research topics in contemporary robotics is to provide robots with a robotic brain that is both...capable (able to solve the assigned taks robustly) and...expandable (it should be possible to add new skills). In the Adaptive systems research group, we are working with a general-purpose structure for robotic brains. In this course, you will use a simplified version of this structure.

14 Programming The robotic brain structure has been implemented in C#, which is also the (main) programming language used in this course. You will need to learn C#. However, you will be given a lot of basic code, so that you can focus on the programming relevant for the course.

15 Course contents (v.3-7) Date Time Room Contents ET Practical robot work ET Practical robot work ET Practical robot work ET Practical robot work ET Practical robot work Obviously, you will need to do a lot work outside the class hours as well. The detailed lab schedule depends on the number of participants, and will be announced during v. 2

16 Assignments Basic robot kinematics (theoretical assignment) You will study the kinematics of the Hubert robot Robot locomotion You will use the Kondo KHR-1 robot (simulated and real) to generate a motion pattern to solve a particular task Text-to-speech (TTS) synthesis Starting from C# classes provided by me, you will write a basic TTS engine (in C#) You will use the TTS engine in connection with the basic robot head, combining TTS, speech recognition, and image processing.

17 Assignments Face tracking and face recognition Using the Hubert robot, you will implement (in C# and PBasic) a robotic brain capable of basic coordinated motion

18 Literature and programming Literature Lecture notes (not entirely complete...) Various web links (e.g. to scientific papers, data sheets, manuals etc.) Programming languages PBasic C#

19 Examination The examination will be based on the assignments and the practical robot work. The detailed requirements (in terms of the number of points on assignments) will be given during v. 2. Assignments will consist of several levels. The basic level must be solved for a passing grade. Solving higher levels makes it possible to receive a higher grade. Some assignments will involve individual work and some will involve group work (group size: 4-5 students)

20 Examination NOTE: In order to get a passing grade, you must also return all robot equipment that you have borrowed. This is an absolute requirement for obtaining the final grade!

21 Work groups The work groups will consist of 4-5 students. Each group will need to come up with a group name! The can be no more than four groups (due to limited availability of robotic hardware).

22 To do (today!) Send me a registration today, containing the following information: Your name Your civic registration number (personnummer) Students who do not send this information today will be assigned (by me) to a group. Note: Preferably, each group should contain at least one student with a laptop (if possible with a serial port (the laptop, not the student ))

23 Download and install...(today!) 1)...C#, either the express version available at or the full version (Microsoft visual studio 2008) available on the program server. 2)...The Basic Stamp Editor (v2.4.2) available at

24 Download (today!) Documentation for C# and PBasic can be found on the web. You start familiarizing yourselves with these languages. Of course, I will provide information regarding both C# and Pbasic, during the last two lectures in v. 2. Of course, you will also (during v.2) be given some sample code, both in C# and PBasic. Note: ALL code that you write MUST follow the code standard, which will be made available on the web page (later this week).

25 Student-teacher interaction During the course you are (of course) very welcome to ask questions at any time, not only during lectures. In order to do so, preferably come to my office (see the directions on the next slide). As a second alternative, send an . However, since I receive many s every day, the best way of asking questions is to come to my office. Note: Please make use of the possibility to ask questions.

26 Office location Mechanical engineering building (nya M-huset), 1st floor. Enter near Café Bulten, follow the blue signs to Applied Mechanics (Tillämpad mekanik) Dial my extension (3727) at the door. My office Bulten Café

27 Introduction to humanoid robots See chapter 1 in the lecture notes...

28 Robot types Stationary robots Manipulators (used in industry) Mobile robots Wheeled Walking (Flying etc.) Note: The distinction between robots and non-robot mechanical systems is not always entirely clear. Example: Automatic braking in cars (just before a collision).

29 Walking robots Monopedal

30 Walking robots Monopedal Bipedal

31 Walking robots Monopedal Bipedal Quadrupedal

32 Walking robots Monopedal Bipedal Quadrupedal Sextupedal etc.

33 Bipedal robots Humanoid robots: Robots with an approximately human-like shape (anthropomorphic) Androids: Robots that (attempt to) exactly look like humans, including skin etc. In some cases (e.g. for HRI) studies a partial humanoid can be sufficient.

34 Why humanoid robots? More natural to interact with (for people), than wheeled robots. Designed (by construction) for operation in environments designed for people. Capable of walking in stairs etc. Engineering challenge.

35 Challenges in humanoid robotics Generating robust, natural (human-like) movements. Extending the time of operation (battery technology ). Development of capable and extendable robotic brains (robotic behaviors, decision-making systems). Reducing the cost of humanoid robots. Legal issues

36 An illustration...

37 Another illustration...