Lecture 04
- 1. Introduction to RoboticsSensorsCSCI 4830/7000September 20, 2010NikolausCorrell
- 2. Review: Kinematics and ControlConceptsForward Kinematics“Odometry”Feed-back ControlInverse Kinematics
- 3. Forward KinematicsHow does the robot move in world space given its actuator speed and geometry?“Odometry”: forward kinematics for mobile platformExample: from exercise 3
- 4. Proportional ControlN.B.: zero error neq correct position!
- 5. More on robot kinematics (arms)John CraigIntroduction to RoboticsMark Spong, Seth Hutchinson and M.VidyasagarRobot Modeling and Control
- 6. Inverse KinematicsHow do we need to control the actuators to reach a certain position?Inversion of forward kinematicsExamples: Differential wheel drive (Exercise 3)
- 7. Feedback controlUse error between reference and actual state to calculate next control inputChange in speed proportional to errorError zero -> speed zeroProblem: find stable controllersExample: from exerciseK. OgataModern Control Engineering
- 8. Today Perception: Basis for reasoning about the worldUnderstand how a sensor works before using itCase studies
- 9. iRobotRoomba4 Bumpers2 Floor sensors1 infrared distance (side)InfraredWheel encoders
- 10. PrairieDogRoomba5.6m, 240 degrees laser scannerIndoor localization systemCameraMicrophone5 Position encoders (arm)
- 11. Nao2 VGA cameras4 Microphones2-axis gyroscope3-axis accelerometer2 bumpers (feet)Tactile sensors (hands + feets)Hall-effect encoders2 Sonar2 InfraredProprioceptive or Exteroceptive?
- 13. Laser Range ScannerMeasures phase-shift of reflected signalExample: f=5MHz -> wavelength 60m
- 14. Examples2 D3D (PR2 sweep)(after classification)
- 15. Sensor performanceDynamic range: lowest and highest readingResolution: minimum difference between valuesLinearity: variation of output as function of inputBandwidth: speed with which measurements are deliveredSensitivity: variation of output change as function of input changeCross-Sensitivity: sensitivity to environmentAccuracy: difference between measured and true valuePrecision: reproducibility of resultsHokuyo URG
- 16. Relation between sensor physics and performance (solutions)Dynamic range: Range: limited by power of light and modulated frequency, smallest wave-length difference measurableAngle: limited by physical setup / trade-off between bandwidth and angular resolutionResolution:Range: Precision of phase-shift measurementAngle: limited by bandwidth / encoderLinearity:Range: phase shift is linear -> signal is linear, but: weak reception makes determination of phase harderAngle: depends on motor implementationBandwidthRange: speed of light, calculating phase shiftAngle: motor speedSensitivity:Range: Doppler effect -> not relevant in robotics, Confidence in the range (phase/time estimate) is inversely proportional to the square of the received signal amplitudeAngle: n.a.Cross-Sensitivity:Range: Glass / reflection properties, 785nm light Accuracy:Range: Precision of phase-shift measurement, strength of reflected lightAngle: motor qualityPrecision: range / variance
- 17. Infra-red distance sensorsPrinciple: measure amount of reflected lightThe closer you get, the more light gets reflectedDigitized with analog-digital converterSharp IR Distance Sensor GP2Y0A02YK20-150cmMiniature IR transceiver0-3cm
- 18. Sensor performanceDynamic range: lowest and highest readingResolution: minimum difference between valuesLinearity: variation of output as function of inputBandwidth: speed with which measurements are deliveredSensitivity: variation of output change as function of input changeCross-Sensitivity: sensitivity to environmentAccuracy: difference between measured and true valuePrecision: reproducibility of resultsSharp IR Distance Sensor
- 19. Relation between sensor physics and performance (solutions)Dynamic range: limited by power of lightResolution: limited by ADC, e.g. 10bit -> 1024 stepsLinearity: highly non-linear (intensity decays quadratically)Bandwidth: limited by ADC bandwidth (sample&hold)Sensitivity: varies over range due to resolutionCross-Sensitivity: sun-light, surface propertiesAccuracy: limited by ADC, varies over rangePrecision: varies over range
- 20. Infra-red distance sensors in Webots (Exercise 1)Color of the bounding object affects sensorNon-linear relation between distance and signal strengthDistance-dependent resolution and noiseSoftware linearizationNoise
- 21. ExerciseDesign a robot that canVacuum a roomMow a lawnCollect golf-balls on a rangeCollect tennis balls on a courtAddressSensorsAlgorithmMechanism
- 22. Scratchboard
- 23. HomeworkRead section 4.1.7 (pages 117 – 145)Questionnaire on CU LearnMidterm: October 11 (during class)