IEEE Robotics & Automation Magazine - March 2022 - 120

EDUCATION
On the Ethics of Research and Practice in
Robotics and Automation
2021 IEEE RAS Winter School on
Simultaneous Localization and
Mapping in Deformable Environments
By Raja Chatila
By Shoudong Huang and Liang Zhao
S
imultaneous localization and
mapping (SLAM) is an important
research problem for robot
navigation in unknown environ -
ments, particularly when GPS is not
available. SLAM requires a robot to
be able to build a map of the environ -
ment in real time and simul taneous -
ly estimate its own location within
the map. In the past two decades,
significant progress has been made in
the research for SLAM in static
environments. However, when an
environment has deformations, such as
when a surgical robot is navigating in
internal body environments, SLAM
needs to build a time-varying 3D map
of the soft tissues and estimate the
location of the robot/sensor within the
map. This poses a challenging problem
since the robot/sensor is moving while
the environment is deforming.
SLAM in deformable environments
is an area of growing interest within the
IEEE Robotics and Automation Society
(RAS) and has the potential to bring
about profound economic and social
benefits. SLAM in deformable environments
has important applications in
many different areas, such as 1) minimally
invasive robotic surgery, 2) animal
body shape reconstruction, 3)
human body motion tracking (e.g., for
sports performance analysis), and 4)
motion capture in virtual reality and
computer games.
The 2021 RAS Winter School on
SLAM in Deformable Environments
Digital Object Identifier 10.1109/MRA.2022.3145919
Date of current version: 22 March 2022
was the first on this topic. The school
aimed to promote education and
research into robot localization, mapping,
and navigation in deformable
environments. It was supported by the
RAS Technical Committee on Computer
and Robot Vision and hosted by
the Robotics Institute at the University
of Technology Sydney (UTS), Australia,
over five days, from 5 to 9 July 2021
(during the southern hemisphere winter).
It was cosponsored by the RAS and
Australia Robotics and Automation
Association. It ran virtually due to
COVID-19 restrictions.
Technical and Social
Program Overview
The school was organized with nine
keynotes, four lectures, four tutorials,
11 group projects, one panel discussion,
and four social networking sessions.
The lectures and tutorials
covered the fundamentals of SLAM,
nonrigid structure from motion,
deformable SLAM, and surgical vision.
Open source SLAM algorithms were
used in the tutorials. We developed
group projects that were closely related
to SLAM in deformable environments.
The projects required the fundamental
algorithms from the lectures and tutorials
and involved practical data sets.
We allocated four 1-h sessions for
social networking, where students
introduced themselves and learned one
another's research interests. On the
final day, we held an award ceremony
and panel discussion about the challenges
and future of SLAM in deformable
environments.
120 * IEEE ROBOTICS & AUTOMATION MAGAZINE * MARCH 2022
Participants
The strong technical program, including
keynotes from leading researchers, and
high-quality lectures/tutorials/projects,
helped attract participants. We also did a
lot of publicity work, such as designing
the school's web page, creating posters,
and publishing advertisements on social
platforms (robotics-worldwide, LinkedIn,
Facebook, Twitter, and WeChat).
The school attracted 228 registrations
from 29 countries in different time
zones. Figure 1 shows the distribution of
the participants and group photos after
the keynotes.
Keynotes, Lectures, and Tutorials
The school invited nine world-leading
robotics scholars from academia and
industry to give keynotes (Figure 2).
The speakers elaborated on the research
of robot localization, mapping, and navigation
in a deformable environment
from different perspectives, including
traditional SLAM, intracorporeal visual
SLAM, how to deform rigid structures,
endoscopic vision research, surgical navigation,
and vision for surgical robotics.
The school included four lectures
and four corresponding tutorials (Figure
3). They disassembled the complex
deformable SLAM problems, starting
from basic theoretical knowledge and
extending it to specific practical applications.
They covered the fundamentals of
SLAM, nonrigid structure from motion,
deformable SLAM, and surgical vision.
To improve the quality of the teaching,
especially for students with little basic
knowledge, the slides were shared in a
OneDrive folder in advance.
IEEE Robotics & Automation Magazine - March 2022

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - March 2022
