Robot Mapping

Introduction to Robot Mapping

Cyrill Stachniss

1
What is Robot Mapping?
§  Robot – a device, that moves through
the environment

§  Mapping – modeling the environment

2
Related Terms

State
Localization
Estimation

Mapping SLAM

Motion
Navigation
Planning
3
What is SLAM?
§  Computing the robot’s pose and the
map of the environment at the same
time

§  Localization: estimating the robot’s

location
§  Mapping: building a map
§  SLAM: building a map and locating
the robot simultaneously
4
Localization Example
§  Estimate the robot’s poses given
landmarks

5
Mapping Example
§  Estimate the landmarks given the
robot’s poses

6
SLAM Example
§  Estimate the robot’s poses and the
landmarks at the same time

7
The SLAM Problem
§  SLAM is a chicken-or-egg problem:
→  a map is needed for localization and
→  a pose estimate is needed for mapping

map

localize
8
SLAM is Relevant
§  It is considered a fundamental
problem for truly autonomous robots
§  SLAM is the basis for most navigation
systems

map

autonomous
navigation

localize
9
SLAM Applications
§  SLAM is central to a range of indoor,
outdoor, in-air and underwater applications
for both manned and autonomous vehicles.

Examples:
§  At home: vacuum cleaner, lawn mower
§  Air: surveillance with unmanned air vehicles
§  Underwater: reef monitoring
§  Underground: exploration of mines
§  Space: terrain mapping for localization
10
SLAM Applications
Indoors Undersea

Space Underground

Courtesy of Evolution Robotics, H. Durrant-Whyte, NASA, S. Thrun 11

SLAM Showcase – Mint

Courtesy of Evolution Robotics (now iRobot) 12

SLAM Showcase – EUROPA

13
Mapping Freiburg CS Campus

14
Definition of the SLAM Problem
Given
§  The robot’s controls

§  Observations

Wanted
§  Map of the environment

§  Path of the robot

15
Probabilistic Approaches
§  Uncertainty in the robot’s motions and
observations
§  Use the probability theory to explicitly
represent the uncertainty

“The robot is “The robot is

exactly here” somewhere here” 16
In Probabilistic Terms
Estimate the robot’s path and the map

distribution path map given observations controls

17
Graphical Model

unknown

observed

unknown

18
Full SLAM vs. Online SLAM
§  Full SLAM estimates the entire path

§  Online SLAM seeks to recover only the

most recent pose

19
Graphical Model of Online SLAM

20
Online SLAM
§  Online SLAM means marginalizing out
the previous poses

§  Integrations are typically done

recursively, one at at time

21
Graphical Model of Online SLAM

22
Why is SLAM a hard problem?
1. Robot path and map are both unknown

2. Map and pose estimates correlated

23
Why is SLAM a hard problem?
§  The mapping between observations and
the map is unknown
§  Picking wrong data associations can have
catastrophic consequences (divergence)

Robot pose
uncertainty

24
Taxonomy of the SLAM Problem
Volumetric vs. feature-based SLAM

Courtesy by E. Nebot 25
Taxonomy of the SLAM Problem
Topologic vs. geometric maps

26
Taxonomy of the SLAM Problem
Known vs. unknown correspondence

27
Taxonomy of the SLAM Problem
Static vs. dynamic environments

28
Taxonomy of the SLAM Problem
Small vs. large uncertainty

29
Taxonomy of the SLAM Problem
Active vs. passive SLAM

Image courtesy by Petter Duvander

30
Taxonomy of the SLAM Problem
Any-time and any-space SLAM

31
Taxonomy of the SLAM Problem
Single-robot vs. multi-robot SLAM

32
Approaches to SLAM
§  Large variety of different SLAM
approaches have been proposed
§  Most robotics conferences dedicate
multiple tracks to SLAM
§  The majority uses probabilistic
concepts
§  History of SLAM dates back to the
mid-eighties

33
SLAM History by Durrant-Whyte
§  1985/86: Smith et al. and Durrant-Whyte
describe geometric uncertainty and
relationships between features or landmarks
§  1986: Discussions at ICRA on how to solve
the SLAM problem followed by the key
paper by Smith, Self and Cheeseman
§  1990-95: Kalman-filter based approaches
§  1995: SLAM acronym coined at ISRR’95
§  1995-1999: Convergence proofs & first
demonstrations of systems
§  2000: Wide interest in SLAM started
34
Three Main Paradigms

Kalman Particle Graph-

filter filter based

35
Motion and Observation Model

"Motion model"

"Observation model"

36
Motion Model
§  The motion model describes the
relative motion of the robot

distribution new pose given old pose control

37
Motion Model Examples
§  Gaussian model

§  Non-Gaussian model

38
Standard Odometry Model
§  Robot moves from to .
§  Odometry information

39
More on Motion Models
§  Course: Introduction to Mobile
Robotics, Chapter 6
§  Thrun et al. “Probabilistic Robotics”,
Chapter 5

40
Observation Model
§  The observation or sensor model
relates measurements with the robot’s
pose

distribution observation given pose

41
Observation Model Examples
§  Gaussian model

§  Non-Gaussian model

42
More on Observation Models
§  Course: Introduction to Mobile
Robotics, Chapter 7
§  Thrun et al. “Probabilistic Robotics”,
Chapter 6

43
Summary
§  Mapping is the task of modeling the
environment
§  Localization means estimating the
robot’s pose
§  SLAM = simultaneous localization and
mapping
§  Full SLAM vs. Online SLAM
§  Rich taxonomy of the SLAM problem

44
Literature
SLAM Overview
§  Springer “Handbook on Robotics”,
Chapter on Simultaneous Localization
and Mapping (1st Ed: Chap. 37.1-37.2)
On motion and observation models
§  Thrun et al. “Probabilistic Robotics”,
Chapters 5 & 6
§  Course: Introduction to Mobile
Robotics, Chapters 6 & 7
45