bridges to computing general information: this document was created for use in the "bridges to...

Download Bridges To Computing General Information: This document was created for use in the "Bridges to Computing" project of Brooklyn College. You are invited

If you can't read please download the document

Upload: michael-nicholson

Post on 14-Dec-2015

218 views

Category:

Documents


0 download

TRANSCRIPT

  • Slide 1

Bridges To Computing General Information: This document was created for use in the "Bridges to Computing" project of Brooklyn College. You are invited and encouraged to use this presentation to promote computer science education in the U.S. and around the world. For more information about the Bridges Program, please visit our website at: http://bridges.brooklyn.cuny.edu/http://bridges.brooklyn.cuny.edu/ Disclaimers: All images in this presentation were created by our Bridges to Computing staff or were found online through open access media sites and are used under the Creative Commons Attribution-Share Alike 3.0 License. If you believe an image in this presentation is in fact copyrighted material, never intended for creative commons use, please contact us at http://bridges.brooklyn.cuny.edu/ so that we can remove it from this presentation.http://bridges.brooklyn.cuny.edu/ This document may include links to sites and documents outside of the "Bridges to Computing" domain. The Bridges Program cannot be held responsible for the content of 3 rd party sources and sites. Slide 2 Introduction to Robotics Lecture 2 Bridges to Computing 2011 M. Meyer Slide 3 Topics: What is a robot? Robot Components Introduction to the course robots (RCX) Programming in Robolab Slide 4 (1) What is a robot? WE will be focusing on AUTONOMOUS mobile robots. These robots can be considered "agents" in the physical world. So... Questions: What is an agent? What is autonomy (autonomous)? What is a robot? Slide 5 (1) Agents Agent is derived from the Latin agere (to do). What an "agent" is depends on the context of the domain. Real estate agent. Secret agent. Software agent, chemical agent. An "agent" is something that "acts" on behalf of some other entity to achieve a goal. Additional Constraint (Russel & Norvig, 32): Agent is "anything that can be viewed as perceiving it's environment through sensors and acting on it's environment through effectors." Slide 6 Example: Software Agent (Game) Slide 7 (1) Autonomy (Autonomous) To be autonomous means: to be independent free of constraints able to act on ones own initiative NO REMOTE CONTROL!!!! NO TELE- OPERATION!!! Autonomous Agent: An agent that while working on behalf of someone else, makes decisions on it's own, guided by feedback from it's sensors. We usually assume autonomy, when discussing agents and robots. Slide 8 (1) Robot definition "A re-programmable, multifunctional manipulator designed to move material, parts, tools, or specialized devices through various programmed motions for the performance of a variety of tasks." Robot Institute of America, 1979. For our purposes a robot is: an autonomous embodied agent. an agent acting independently, whose environment is the real world! We ASSUME autonomy when we use the term robot. Slide 9 (1) Our definition of a robot A robot is an autonomous system (agent) which exists in the physical world, that can sense its environment (including its own internal state) and act on its environment to achieve some goals. Robotics: The study of robots: their design, construction, capabilities and purpose. Slide 10 (2) Types of Robotic Control 1. Reactive Control 2. Hybrid Control 3. Behavior Based We will go into these topics in detail later in the semester. But here is a question to get you started. What kind of control system would you prefer if you are in a room with: 1. A rapidly spreading fire. 2. Several land-mines. Slide 11 (3) Robot Components Our robot definition: "A robot is an autonomous system, which exists in the physical world, that can sense it's environment, and can act on it to achieve some goals." So our robot must have: 1.A body 2.Sensors 3.Effectors 4.A controller Slide 12 (3) Embodiment In a software environment (ex: a game world) we get to make the rules, and we can let an agent do anything we want. In the real world: 1. The laws of physics apply (gravity, friction, entropy). 2. Objects can't overlap (collision avoidance). 3. Physical bodies have range, strength, distance and shape limitations. 4. Physical bodies have time limitations. Slide 13 (3) Sensing A robot gathers information about its state and the environment via sensors. Sensors can be classified as active and passive. Typically a robot has a suite of sensors capable of monitoring numerous features, such as battery level, odometry, and distance to nearby objects. With this information, the robot can determine its current state. The robot can then use this state information to decide what actions are appropriate. In the end, all sensors, are converting a physical property, into an electronic signal. Slide 14 (3) Types of Sensors Property being sensedType of sensor contactbump, switch distanceultrasound, radar, infra-red, laser light levelphoto cell, camera sound levelmicrophone smellchemical temperaturethermal inclinationgyroscope rotationencoder pressurepressure gage altitudealtimeter Slide 15 (3) State 1. May be hidden, partially observable, observable (don't have perfect knowledge in the real world). 2. May be discrete (0/1) or continuous (3.33 m/sec). 3. State space refers to all of the possible values (may be infinite) that a systems state could be in. 4. We can conceptualize both an internal and external state space. 5. Robots may carry around a representation (model) of the external world, as part of their internal state. Slide 16 (3) Action 1. Effectors enable a robot to take action, to change the state of the world (including its own position). 2. Actuators are the underlying mechanisms (muscles, motors, solenoids) which do the actual work. 3. Main action activities are: o Locomotion (moving around) o Manipulation (handling objects) 4. Degrees of freedom: Refers to the range of motion, the dimensions in which a manipulator can move (more on this as well as Pitch, Yaw and Roll in Chap 6). Slide 17 (3) Controllers Robots utilize small, highly specialized computers to allow them to reason about their world. The kind of reasoning employed varies (reflexive versus "intelligence"). We can talk about two kinds of intelligence in our controllers. Reasoning (if/then) Learning (this is what the past was like... so...) Slide 18 (4) Say Hello to the RCX The Robotics Command Explorer is the brain of the MINDSTORM. The RCX is a small computer (nickname "programmable brick"). It has: o An IR transceiver o 3 input ports (1-3) light sensors bumpers o 3 output ports (A-C) motors lights Slide 19 (3) For the Hardware Junkies Slide 20 (3) Programming You will write your programs on a computer and download them to the RCX using an IR transmitter ("communications tower"). We will use Robolab a VPL (Visual Programming Language) IDE (integrated development environment) to write our programs. There are other program interfaces to the RCX. Slide 21 End of Part II Unit A Lecture You are now assigned your second homework: Assessment Assignment #2 - Big Dog Read 2 case studies. Answer some short questions It is due at the start of UNIT B. No late assignments will be accepted. Slide 22 -------- Side note: -------- Halfway through this lecture, we looked at some of the NEGATIVE aspects of the coming "robot revolution". Which leads us to the question: Do we even want robots? The answer is (at least for me) yes. Why? Because there are some jobs, some tasks, that we don't want human beings to have to do, because they are: 1. DULL 2. DIRTY 3. DANGEROUS Slide 23 -------- Side note: -------- DULL Slide 24 -------- Side note: -------- DIRTY Slide 25 -------- Side note: -------- DANGEROU S Slide 26 -------- Side note: -------- Regarding Military and Police robots. What would a robot be programmed to do in the following situations? 1. Confronting a suspect in a dark doorway who is pointing a black object at the robot. 2. Apprehending a suspect after a dangerous high-speed car chase through city streets. 3. Confronting student protesters on a college campus, one of whom has just hurled a bottle. In a crisis situation, human beings often cease to be rational beings and instead act based on their training, emotions and instincts. ("On Killing" by LT. Colonel Dave Grossman). Consider: A robot won't care if you spit on it. Slide 27 -------- Side note: -------- Big picture: Robots and robotics pose both huge problems and huge opportunities for all mankind. The question is, can we, as a species, adapt to this new and disruptive technology. Something fun to ponder: In the end, the earth, our planet and home, will be destroyed by the Sun. We are not built to survive/thrive in outer space. There are some who have suggested that Robots may well be the next step in our evolution...