Autonomous Mobile Robots

ETH Master Course: 151-0854-00L

Autonomous Mobile
Robots
Lecture: Monday 14.15 - 16.00, HG D 3.2
Exercises: Monday 16.15 - 18.00, HG K32 (VisDome)
(ONLY lecture of 26 April, 14.15 – 18.00, HG D2.1)

Roland Siegwart
Davide Scaramuzza

Zürich Autonomous Systems Lab

1
2 “Robots” Getting Closer
 Technical systems can be characterized
by the increasing physical and
psychological closeness and interaction Industrial Robots
between man and machine

Service and
Personal Robots

Cyborgs
Exponential increase of complexity
© R. Siegwart, ETH Zurich - ASL
 1 - Introduction
1
3 Autonomous Mobile Robots
 The three key questions in Mobile Robotics
 Where am I ?
 Where am I going ? ?
 How do I get there ?

 To answer these questions the robot has to

 have a model of the environment (given or autonomously built)
 perceive and analyze the environment
 find its position within the environment
 plan and execute the movement
 This course will deal with Locomotion and Navigation (Perception,
Localization, Planning and Motion Generation)

© R. Siegwart, ETH Zurich - ASL

1
4 ASL – ETH Zurich
 Micro Air Vehicles

 Walking and Running

Quadruped Robots

 Service Robots

 Autonomous Robots/Cars
for Inner City Environments

 Inspection Robots

 Space Robots for Planetary

Exploration

 Swimming Robots

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
5 Content of the Course
1. Introduction

2. Locomotion

3. Mobile Robot Kinematics

4. Perception

5. Mobile Robot Localization and Mapping

6. Planning and Navigation

 Other Aspects of Autonomous Mobile Systems

 Applications

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
6 Slides and Lecture Notes
 Slides and Exercises
 http://www.asl.ethz.ch/education/master/mobile_robotics
 Currently you can find there the last years slides and exercises. They will be updated
soon.

 Lecture Notes:
 Introduction to Autonomous Mobile Robots
Roland Siegwart, Illah Nourbakhsh, Davide Scaramuzza
• Intelligent Robotics and Autonomous Agents series
• The MIT Press
• Massachusetts Institute of Technology
• Cambridge, Massachusetts 02142
• ISBN 0-262-19502-X
 http://www.mobilerobots.org

 Can be purchased in our secretariat (CLA E31)

for 40 CHF.

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
7 Program

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
8 From Manipulators to Mobile Robots

© R. Siegwart, ETH Zurich - ASL

1
9 “Seeing” the world – more than appearance
 Perception and models (“understanding”) are strongly linked

What is the
difference
in brightness?

http://web.mit.edu/persci/people/adelson/checkershadow_downloads.html
© R. Siegwart, ETH Zurich - ASL
 1
10 “Connecting” with the world
 Tactility, key for controlling the real world

Courtesy of DLR, Germany

Video courtesy of DLR, Germany

 It takes us around 14 years

to learn holding a glass with an optimal force
© R. Siegwart, ETH Zurich - ASL
 1 - Introduction
1
11 General Control Scheme for Mobile Robot Systems

Knowledge, Mission
Data Base Commands

Localization Cognition
"Position"
Map Building Path Planning
Global Map

Environment Model Path

Local Map

Information Path
Extraction Execution

Motion Control
Perception

Raw data Actuator Commands

Sensing Acting

Real World
Environment

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
12 Applications of Mobile Robots
Indoor Outdoor
Structured Environments Unstructured Environments

mining
transportation
industry & service space sewage tubes

customer support agriculture

museums, shops .. forest
cleaning ..
air
large buildings
research, construction
entertainment, surveillance demining
toy buildings underwater
fire fighting
military

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
13 Automatic Guided Vehicles

 Newest generation of
Automatic Guided Vehicle
of VOLVO used to
transport motor blocks
from on assembly station
to an other. It is guided by
an electrical wire installed
in the floor but it is also
able to leave the wire to
avoid obstacles. There are
over 4000 AGV only at
VOLVO‟s plants.

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
14 Helpmate

 HELPMATE is a mobile robot used in hospitals

for transportation tasks. It has various on board
sensors for autonomous navigation in the
corridors. The main sensor for localization is a
camera looking to the ceiling. It can detect the
lamps on the ceiling as reference (landmark).
http://www.ntplx.net/~helpmate/

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
15 BR700 Cleaning Robot

 BR 700 cleaning robot

developed and sold by
Kärcher Inc., Germany.
Its navigation system is
based on a very
sophisticated sonar
system and a gyro.
http://www.kaercher.de

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
16 ROV Tiburon Underwater Robot

 Picture of robot ROV Tiburon for

underwater archaeology (teleoperated)-
used by MBARI for deep-sea research,
this UAV provides autonomous hovering
capabilities for the human operator.

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
17 The Pioneer

 Picture of Pioneer, the teleoperated robot that is

supposed to explore the Sarcophagus at Chernobyl

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
18 Forester Robot

 Pulstech developed the

first „industrial like‟ walking
robot. It is designed
moving wood out of the
forest. The leg
coordination is automated,
but navigation is still done
by the human operator on
the robot.
http://www.plustech.fi/

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
19 Sojourner, First Robot on Mars

 The mobile robot

Sojourner was used
during the Pathfinder
mission to explore the
mars in summer 1997.
It was nearly fully
teleoperated from
earth. However, some
on board sensors
allowed for obstacle
detection.
http://ranier.oact.hq.na
sa.gov/telerobotics_pa
ge/telerobotics.shtm

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
20 The Honda Walking Robot http://www.honda.co.jp/tech/other/robot.html

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
21 General Control Scheme for Mobile Robot Systems

Knowledge, Mission
Data Base Commands

Localization Cognition
"Position"
Map Building Path Planning
Global Map

Environment Model Path

Local Map

Information Path
Extraction Execution

Motion Control
Perception

Raw data Actuator Commands

Sensing Acting

Real World
Environment

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
22 Control Architectures / Strategies

 Control Loop
 dynamically changing  Two Approaches
 no compact model available  Classical AI
• complete modeling
 many sources of uncertainty
• function based
• horizontal decomposition

"Position"
 New AI, AL
Localization Cognition
Global Map • sparse or no modeling
• behavior based
Environment Model Path
Local Map • vertical decomposition
• bottom up
Perception Real World Motion Control
Environment

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
23 Two Approaches

 Classical AI
(model based navigation)
 complete modeling
 function based
 horizontal
decomposition
 New AI, AL
(behavior based navigation)
 sparse or no modeling
 behavior based
 vertical decomposition
 bottom up
 Possible Solution
 Combine Approaches

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
Environment Representation and Modeling:
24 The Key for Autonomous Navigation

 Environment Representation
 Continuos Metric -> x,y,
 Discrete Metric -> metric grid
 Discrete Topological -> topological grid

 Environment Modeling
 Raw sensor data, e.g. laser range data, grayscale images
• large volume of data, low distinctiveness
• makes use of all acquired information
 Low level features, e.g. line other geometric features
• medium volume of data, average distinctiveness
• filters out the useful information, still ambiguities
 High level features, e.g. doors, a car, the Eiffel tower
• low volume of data, high distinctiveness
• filters out the useful information, few/no ambiguities, not enough information

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
25 Environment Representation: The Map Categories

 Recognizable Locations  Topological Maps

Courtesy K. Arras
 Metric Topological Maps  Fully Metric Maps (continuos or
discrete)
y

200 m
50 km

2 km

100 km

x
{W}
© R. Siegwart, ETH Zurich - ASL
 1 - Introduction
1
26 Methods for Navigation: Approaches with Limitations

 Incrementally  Modifying the environments

(dead reckoning) (artificial landmarks / beacons)

Courtesy K. Arras
Odometric or initial sensors (gyro)

Inductive or optical tracks (AGV)

Reflectors or bar codes

 not applicable
 expensive, inflexible

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
27 Gaining Information through motion: (Multi-hypotheses tracking)

Believe state

Courtesy S. Thrun, W. Burgard

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
28 Methods for Localization: The Quantitative Metric Approach
1. A priori Map: Graph, metric
3. Matching:
y
Find correspondence
of features

Courtesy K. Arras
wy lw
r r

{W} wx
r x
4. Position Estimation:
2. Feature Extraction (e.g. line segments) e.g. Kalman filter, Markov

Odometry Observation

 representation of uncertainties
 optimal weighting acc. to a priori statistics

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
29 Map Building: The Problems

1. Map Maintaining: Keeping track of 2. Representation and

changes in the environment Reduction of Uncertainty

Courtesy K. Arras
e.g. disappearing position of robot -> position of wall
cupboard

position of wall -> position of robot

- e.g. measure of belief of each  probability densities for feature positions

environment feature  additional exploration strategies
© R. Siegwart, ETH Zurich - ASL
 1 - Introduction
1
30 High-Speed Exploration and Mapping

Courtesy of Sebastian Thrun

© R. Siegwart, ETH Zurich - ASL
 1 - Introduction
1
31 Tour-Guide Robot (EPFL @ expo.02)

© R. Siegwart, ETH Zurich - ASL

 1 - Introduction
1
32
DARPA Grand Challenge
 Hundreds of miles autonomouly through the desert

© R. Siegwart, ETH Zurich - ASL