IEEE Robotics & Automation Magazine - June 2021 - 121

IOC-AUDECO relied on Aruco markers to enhance the
robustness of the table localization. Robotics.SG wrapped paper
around the shelf legs to facilitate the mapping, navigation, and
localization of the mobile base.
We were also interested in finding out if the experience
gained during the hackathon was exploited afterward. Of
the four teams that provided an answer to this question,
three indicated that they evolved some of the components
from the demonstration to create a more advanced lab version
(IOC-AUDECO and TAMS) or reused solutions for
a new competition (Homer). The TAMS demonstration, for
instance, was transferred to a different platform, a PR 2 robot,
thus showing the generality of the team's solution (https://
www.youtube.com/watch?v=8S2MvKNbwmM). Three
teams (IRI, IOC-AUDECO, and TAMS) indicated that some
of the components were further developed and already published
or submitted for publication as scientific papers. IRI
has been able to transfer the knowledge gained from the
force loop controller in the feeding task to a scenario involving
bimanual cloth manipulation [43]. IOC-AUDECO continued
to develop task and motion planning for mobile
manipulation executions [44]. TAMS advanced object perception
from the competition to detect and reconstruct
transparent items [45].
Finally, we asked the teams to tell us the most important
lesson they learned from the hackathon. IRI highlighted the
need for further supervision during the demonstration execution.
Its members reported that, as a lesson learned, their current
exhibitions are now carefully designed to accommodate
double-check control at different levels. In this line, IOCAUDECO
identified the need for more robust error detection
and recovery strategies to resume tasks and recover from
unexpected situations. TAMS highlighted the benefits of integrating
independent components in a unified demonstration
and recognized the need for intensively testing each component
beforehand to avoid the more difficult job of debugging
the overall execution. Team Homer appreciated the benefit of
having on-site robotic platforms for implementing the demonstration
outside its lab, thus reducing
funding needs and transportation/
insurance costs. Also, its members
noted the benefits of having a common
robotic platform for increasing
the comparability of results across
multiple research groups.
Discussion and Outlook
In this section, we discuss the lessons
learned after organizing the hackathon
and the outlook for similar
future events.
Hackathon Structure
The Mobile Manipulation Hackathon
challenged the community to show
integrated demonstrations
that
exploited the benefits of a mobile robotic manipulation platform.
This required the development and integration of components
at different levels, e.g., perception, navigation, and
localization; grasp and manipulation planning; and human-
robot interaction. Teams were free to propose a demonstration
script, and they used this opportunity to showcase their
latest developments in those fields, as opposed to a fixed task.
This was a key difference of our hackathon when compared
with competitions that have predetermined challenges. We
believe both types of events are beneficial to the field: competitions
with a fixed task provide a clearer picture of progress in
a particular area, while competitions with an open structure,
such as ours, are useful to understand a variety of possible
applications. Therefore, we encourage the community (and
we intend) to organize both types in the future.
Use of a Fixed Demonstrator Platform
The opportunity to use a unified hardware/software platform
based on ROS provided a chance to compare multiple
approaches. A solid software and simulation framework
enabled teams to remotely develop their demonstration, thus
reducing the time required for physical integration in the
robotic system. However, we recognize that the basic tools for
fast prototyping and quick debugging still need to be
enhanced to facilitate the integration of full systems within a
few days. In terms of the competition, it was greatly beneficial
to have the robots on site, thus relieving teams of the burden
of worrying about transportation costs, insurance, basic setups,
and infrastructure and enabling them to focus on the
pure development process.
From the perspective of robot manufacturers, the hackathon
was a great opportunity to gather valuable feedback
from both experienced and new users of the robots, which
will help to improve how the next generation of devices will
be conceived. The research community can also benefit from
this kind of competitions to identify tools, libraries, and
frameworks that could accelerate the implementation of realworld
applications with complex robots, such as mobile
Object Detection
Localization of Mobile Base
Navigation and Mapping
Wi-Fi Communication
Overall Software Integration
Hand-Eye Calibration
Figure 6. Challenging areas in the hackathon demonstrations.
JUNE 2021 * IEEE ROBOTICS & AUTOMATION MAGAZINE *
121
12 34 5
http://www.Robotics.SG http://www.youtube.com/watch?v=8S2MvKNbwmM http://www.youtube.com/watch?v=8S2MvKNbwmM
IEEE Robotics & Automation Magazine - June 2021

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - June 2021
