Date post: | 11-Apr-2018 |
Category: |
Documents |
Upload: | nguyenphuc |
View: | 226 times |
Download: | 4 times |
1Artificial Intelligence - RoboTutor
Nao by NaoAn introduction to robotics
2Artificial Intelligence - RoboTutor
Overview
•Introduction to Robotics
•Sensors of NAO
•Effectors of NAO
•Programming a NAO
•Applications – Robotutor
3Artificial Intelligence - RoboTutor
The word “robot” is derived from a Slavonic word which means serf.
A. TrueB. False
True
False
0%0%
9
4Artificial Intelligence - RoboTutor
My (distant) robot family
Asimo
KUKA
DaVinci
GeminoidManipulators Mobile Robots
Zebro
Quadcopter
Humanoids
5Artificial Intelligence - RoboTutor
Why do you need us robots?
Tasks dangerous for humans Repetitive, mechanical tasks
Service tasksExploration tasks
6Artificial Intelligence - RoboTutor
What is a robot?
A robot is:
• Physical Agent
• Manipulates the physical world
• Has sensors
• Has effectors
7Artificial Intelligence - RoboTutor
Which of these, according to the definition, is not a robot?
A. ZebroB. A classic vacuum
cleanerC. Cruise controlD. The Mars rover
A. B. C. D.
0% 0%0%0%
9
8Artificial Intelligence - RoboTutor
How do we work?
Sensor
Control Program
Effector
9Artificial Intelligence - RoboTutor
Humanoid Robot NAO
• Humanlike appearance
• Programmable, compact, electro-
mechanical
• Processor: Intel Atom Z530
• Software
• Linux System
• Control Architecture: NaoQi Framework
• (AL)Modules and self-built modules
• Multi-language SDK
10Artificial Intelligence - RoboTutor
Remote and Local Control
ServerClient
Local vs. Remote
Trade-off• Computational power• Data transmission cost
Atom Z530 on the robot
Core i7 on the desktop/laptop
11Artificial Intelligence - RoboTutor
Sensors
• Sensors• Perceive properties and events
from environment• Color, shape, size of objects• A human is approaching the robot
• Feedback on its own behaviors• Is the robot hand touching the cup?
• Example sensors:• Gyroscope, thermometer, laser sensor, camera, microphone, etc.
lasers
infraredcamera
digital compass
sonar
photoelectricencoder
GPS
12Artificial Intelligence - RoboTutor
Sensors of NAO
• Camera 2• Microphone 4• Sonar 2• Pressure 8• Tactile 9• Bumper 4• Button 1• Infrared 2• Accelerometer 1• Gyro 2• Joint Position Sensor 36• Current/Temperature
Resistors 27
13Artificial Intelligence - RoboTutor
Sensor – Position sensors(Magnetic Rotary Encoder)
• Detect the (angular) position of
joints/motors
• In the picture
• Two equally spaced magnets
• The output voltage peaks twice
per revolution.
14Artificial Intelligence - RoboTutor
Force sensor- Force Sensitive Resistors
• Detect whether the robot is standing on something
• Force change ->Resistor value change
• Walking balance
• Fall-down detection
15Artificial Intelligence - RoboTutor
Tactile Sensors• Hand (3)
• Left, Right, Back
• Head (3)• Front, middle, back
• Foot (2x2)• Left, right
16Artificial Intelligence - RoboTutor
Inertial and Gyro sensors
• Acceleration: x, y, z• Rotation: Roll (x), Pitch (y)
• Tracking the robot position
• Tracking the robot pose
• Walking balance
• Fall-down detection
17Artificial Intelligence - RoboTutor
Sensor - Sonar
• Emitter-receiver pairs ×2 on front
• Resolution: 1cm
• Effective cone: 60°
• Detection Range: 0.25m to 2.55m
• Obstacle detection
• Pros: Fast
SonarRightNothingDetected
SonarLeftNothingDetected
SonarRightDetected
SonarLeftDetectedEvents:
18Artificial Intelligence - RoboTutor
Sensor - Microphone
• 4: front, rear, left, right
• Speech recognition• Match predefined words • Cannot recognize undefined
words
• Audio source localization
19Artificial Intelligence - RoboTutor
Sensor - Camera
• Number×2 on front
• Active Pixels 1288×968
• Frame Rate
• Non-overlapped field of View
• Focus Range: 30cm ~ infinity
20Artificial Intelligence - RoboTutor
Traditional stereo vision cannot be constructed directly using the cameras on the NAO because
A. They are not located horizontally
B. They have no overlapped field of view
C. Computational power is insufficient on the robot
D. Software for stereo-vision is not available for the NAO
A. B. C. D.
0% 0%0%0%
9
21Artificial Intelligence - RoboTutor
How many sensors do you think I have?
A. 20 – 40B. 40 – 60C. 60 – 80 D. 80 – 100
20 –
40
40 –
60
60 –
80
80 –
100
0% 0%0%0%
9
22Artificial Intelligence - RoboTutor
Which module is the most suitable to run on the robot?
A. Sonar based obstacle avoidance
B. 3D reconstruction of environment based on computer vision
C. Speech-to-text recognition
D. Facial expression recognition
A. B. C. D.
0% 0%0%0%
15
23Artificial Intelligence - RoboTutor
Break