MECHANICAL
ENGINEERING
ME 652
Mobile Robotics
Handout #1:
Course Introduction
Spring, 2020
Jinwhan Kim
KAIST
MECHANICAL
ENGINEERING
Course Information
• Course instructor
– Jinwhan Kim ([email protected])
– Office: C323, KI Building (E4)
– Office hours: Appointment-basis via email
• Teaching assistants
– Dasol Jeong ([email protected])
– Junwoo Jang ([email protected])
• Prerequisite
– Familiarity with the basic (undergraduate-level) linear
algebra and probability theory
– Prior exposure to MATLAB is a plus but not a must
ME652 by Jinwhan Kim, KAIST ME 2
MECHANICAL
ENGINEERING
Grading Policy
• Attendance (10%)
• Homework (30%)
– 5 programming assignments
– Note: Late submission will be penalized 20% per
day after the due date.
• Quiz (30%)
– Open-book, in-class
• Term project (30%)
ME652 by Jinwhan Kim, KAIST ME 3
MECHANICAL
ENGINEERING
Term Project
• The term project can be done in groups of 2-4
persons.
– More work is expected from larger groups.
• Project evaluation
– Three in-class presentations by students
• Proposal, milestone & Final
– Final written report
• The report should not exceed 10 pages.
ME652 by Jinwhan Kim, KAIST ME 4
MECHANICAL
ENGINEERING
Topics to be covered
• Topics to be covered in this course
– Basic kinematics & dynamics of robotic vehicles
– Control theory and vehicle guidance laws
– Tree/Graph search algorithms
– Single and cooperative motion planning
– Bayesian estimation, Kalman filter, particle filter
– Vehicle navigation, mapping algorithms and SLAM
– Machine learning, reinforcement learning
ME652 by Jinwhan Kim, KAIST ME 5
MECHANICAL
ENGINEERING
Course Schedule
• Schedule
ME652 by Jinwhan Kim, KAIST ME 6
MECHANICAL
ENGINEERING
Course References
• Siegwart, Nourbakhsh and Scaramuzza, Introduction to
Autonomous Mobile Robots, 2nd Ed., MIT Press, 2011.
• Thrun, Burgard and Fox, Probabilistic Robotics, MIT Press, 2005.
• Choset et al., Principles of Robot Motion, MIT Press, 2005.
• Fossen, Guidance and Control of Ocean Vehicles, John Wiley &
Sons, 1994.
• Nelson, Flight Stability and Control, 2nd Ed., McGraw-Hill, 1997
• Russel and Norvig, Artificial Intelligence: A Modern Approach,
3rd Ed., Prentice Hall, 2009
• Sutton and Barto, Reinforcement Learning: An Introduction, 2nd
Ed., Bradford Book, 2018
ME652 by Jinwhan Kim, KAIST ME 7
MECHANICAL
ENGINEERING
What is robot?
• A machine capable of carrying out a complex
series of actions automatically, especially one
programmable by a computer (Oxford dictionary).
cf.: The word “robot” derives from the Czeck word for "worker”
iRobot? iRobot!
[Videos from YouTube]
ME652 by Jinwhan Kim, KAIST ME 8
MECHANICAL
ENGINEERING
Aerial Drones & Self-Driving Cars
[Images from Google]
ME652 by Jinwhan Kim, KAIST ME 9
MECHANICAL
ENGINEERING
Indoor Navigation with a UAV
https://www.youtube.com/watch?v=IMSozUpFFkU
ME652 by Jinwhan Kim, KAIST ME 10
MECHANICAL
ENGINEERING
Wall Climbing Robot
https://www.youtube.com/watch?v=YOk5pMwY5Co
ME652 by Jinwhan Kim, KAIST ME 11
MECHANICAL
ENGINEERING
Robot Collaboration in Warehouse
https://www.youtube.com/watch?v=FBl4Y55V2Z4
ME652 by Jinwhan Kim, KAIST ME 12
MECHANICAL
ENGINEERING
Navigation & Mapping on Campus
ME652 by Jinwhan Kim, KAIST ME 13
MECHANICAL
ENGINEERING
3D Reconstruction
ME652 by Jinwhan Kim, KAIST ME 14
MECHANICAL
ENGINEERING
Surface & Underwater Mapping
ME652 by Jinwhan Kim, KAIST ME 15
MECHANICAL
ENGINEERING
Vision-Based Navigation
ME652 by Jinwhan Kim, KAIST ME 16
MECHANICAL
ENGINEERING
In-Water Visual Hull Inspection
Depth: 1m Depth: 4m Depth:5m Depth:7.5m
Drillship
West Draco
built by SHI
AUV model with the camera
footprint (red) and vehicle
trajectory (blue)
ME652 by Jinwhan Kim, KAIST ME 17
MECHANICAL
ENGINEERING
In-Water Visual Hull Inspection
Start
ME652 by Jinwhan Kim, KAIST ME 18
MECHANICAL
ENGINEERING
Robot Classifications
• Classification by applications
– Industrial robots, military robots, service robots, medical
robots, entertainment robots, etc.
• Classification by locomotion
– Stationary robots
– Mobile robots
(or robotic vehicles)
• Wheeled robots
• Legged robots
• Swimming robots
• Flying robots
• Others
[Images from Google]
ME652 by Jinwhan Kim, KAIST ME 19
MECHANICAL
ENGINEERING
Vehicles
• Vehicle:
“A device that is designed or used
to transport people or cargo.” (wiki)
– Space vehicles
• Spacecraft, Satellites
– Aerial vehicles
• Airplanes, rotorcraft, missiles
– Ground vehicles
• Wheeled cars, legged robots
– Ocean vehicles
• Surface ships, underwater vehicles
[Images from Google]
ME652 by Jinwhan Kim, KAIST ME 20
MECHANICAL
ENGINEERING
Unmanned Robotic Vehicles
• Unmanned Aerial Vehicles (UAVs)
• Unmanned Ground Vehicles (UGVs)
• Unmanned Surface Vessels (USVs)
• Unmanned Underwater Vehicles (UUVs)
[Images from Google]
ME652 by Jinwhan Kim, KAIST ME 21
MECHANICAL
ENGINEERING
Robotic Vehicles
• Robotic vehicles are machines that move “autonomously” on
the ground, in the air, undersea, or in space.
• Robotic vehicles are “unmanned”, in the sense that no
humans are on board, and they move by themselves, under
their own power, with sensors and computational resources
onboard to guide their motion.
• However, such “unmanned” robotic vehicles usually integrate
some form of human oversight or supervision of the motion
and task execution, and such oversight may take different
forms depending on the environment and application.
WTEC Panel Report on INTERNATIONAL ASSESSMENT OF RESEARCH AND DEVELOPMENT IN ROBOTICS, 2006.
ME652 by Jinwhan Kim, KAIST ME 22
MECHANICAL
ENGINEERING
Vehicle Autonomy
Autonomy Levels for Unmanned Systems Framework, Volume I: Terminology, Version 1.1, Huang, H. Ed., NIST Special Publication
1011, National Institute of Standards and Technology, Gaithersburg, MD, September 2004.
ME652 by Jinwhan Kim, KAIST ME 23
MECHANICAL
ENGINEERING
Vehicle Autonomy
• Autonomy levels for unmanned systems
Autonomy Levels for Unmanned Systems Framework, Volume I: Terminology, Version 1.1, Huang, H. Ed., NIST Special Publication
1011, National Institute of Standards and Technology, Gaithersburg, MD, September 2004.
ME652 by Jinwhan Kim, KAIST ME 24
MECHANICAL
ENGINEERING
Degrees of Autonomy
• Remote control
– The human operator, without sensory feedback, directly controls the
actuators of the system. In this mode, the system takes no initiative
and relies on continuous input from the user.
• Tele-operation
– The human operator, using video feedback and/or other sensor
feedback, either directly controls the actuators or assigns incremental
goals. In this mode, the system may take limited initiative in reaching
the assigned incremental goals.
• Semi-autonomous
– An unmanned system and the human operator conduct a mission and
requires various levels of human-robot interactions.
• Fully autonomous
– An unmanned system is expected to accomplish its mission, within a
defined scope, without human intervention.
ME652 by Jinwhan Kim, KAIST ME 25
MECHANICAL
ENGINEERING
Remote Control
[Images from Google]
ME652 by Jinwhan Kim, KAIST ME 26
MECHANICAL
ENGINEERING
Tele-Operation
[Images from Google]
ME652 by Jinwhan Kim, KAIST ME 27
MECHANICAL
ENGINEERING
Semi-Autonomous
[Images from Google]
ME652 by Jinwhan Kim, KAIST ME 28
MECHANICAL
ENGINEERING
Fully Autonomous
• Rule #1: Gain information about the
environment
• Rule #2: Work for an extended period
without human intervention
• Rule #3: Move either all or part of
itself throughout its operating
environment without human
assistance
• Rule #4: Avoid situations that are
harmful to people, property, or itself
unless those are part of its design
specifications
http://en.wikipedia.org/wiki/Autonomous_robot
[Images from Google]
ME652 by Jinwhan Kim, KAIST ME 29
MECHANICAL
ENGINEERING
Robot Intelligence
• Intelligence is required for robots to be able to
achieve a given task autonomously.
• Robots are manufactured as hardware. Artificial
intelligence is a theory based on which the software
for the robots is realized.
• The control of the robot is a software agent that
reads data from the sensors (or perceive
environment), decides what to do next and then
directs the effectors of the robots to act in the
physical world.
ME652 by Jinwhan Kim, KAIST ME 30
MECHANICAL
ENGINEERING
Robot Operation
Learning Knowledge
Database
?!
Perception:
- Sensing world environment
Agent
행동 인식 Planning:
- Computing a sequence of actions
Action:
Environment - Performing pre-computed actions
ME652 by Jinwhan Kim, KAIST ME 31
MECHANICAL
ENGINEERING
Control Block Diagram
Feedforward Path Process Noise
+
u Vehicle x
Guidance + Controller + w/ Actuator
-
^x z
Feedback Path
Estimator Sensor
Measurement Noise
ME652 by Jinwhan Kim, KAIST ME 32
