Probabilistic Algorithms
for
Mobile Robot Mapping
"Based on the paper"
Slide 3
Museum Tour-Guide Robots
The Nursebot Initiative
Slide 6
Mapping: The Problem
Mapping: The Problem
Milestone Approaches
3D Mapping
Take-Home Message
Slide 12
Bayes Filters
Bayes Filters in
Localization
Bayes Filters for Mapping
Kalman Filters (SLAM)
Underwater Mapping with
SLAM
Courtesy of Hugh Durrant-Whyte, Univ of Sydney
Large-Scale SLAM
Mapping
Courtesy of John Leonard, MIT
SLAM: Limitations
Slide 20
Unknown Data Association:
EM
CMU’s Wean Hall (80 x 25
meters)
EM Mapping, Example
(width 45 m)
EM Mapping: Limitations
Slide 25
The Goal
Real-Time Approximation (ICRA
paper)
Incremental ML: Not A
Good Idea
Real-Time Approximation
Real-Time Approximation
Importance of Posterior
Pose Estimate
Online Mapping with
Posterior
Courtesy of Kurt Konolige, SRI, DARPA-TMR
Accuracy: “The Tech”
Museum, San Jose
Multi-Robot Mapping
Multi-Robot Exploration
3D Volumetric Mapping
3D Structure Mapping
3D Texture Mapping
Fine-Grained
Structure:
Can We Do Better?
Slide 40
Multi-Planar 3D Mapping
3D Multi-Plane Mapping
Problem
Expected Log-Likelihood
Function
EM To The Rescue!
Results
The Obvious Next Step
Underwater Mapping (with University of Sydney)
Slide 48
Take-Home Message
Open Problems
Open Problems, con’t
Slide 52