This paper describes Team Tidyboy's approach to build a functional, robust, yet highly modular software framework for home service robots. Home service tasks require a number of software modules to work together, which includes verbal and nonverbal human robot interactions, 3D perception of known and unknown objects and manipulation of various objects using movement of both the manipulator and the base of the robot. World Robot Summit (WRS) 2020 Partner Robot Challenge (PRC) rule has raised the bar higher by asking for a reliable, long-term operation of such a system without human intervention. We present how we have overcome such challenges by robust software modules, repeated experiments in randomly initialized simulated environments, and finally introduction of various failure detection and recovery algorithms. Our software is successfully validated in RoboCup@Home 2021 and WRS 2020 PRC remote participation league, which led us to obtaining the first places in both competitions.
Bibliographical noteFunding Information:
This work was partially funded by Police-Lab 2.0 Program (www.kipot.or.kr) funded by the Ministry of Science and ICT, South Korea (MSIT) and KNPA (082021D48000000), IITP grant funded by (2021-0-01202), KIAT grant funded by the MOTIE (P0008473), IITP (2015-0-00310-SW.Star-Lab/25%, 2019-0-01371-BabyMind/25%, 2021-0-02068-AIHub/25%) grants, and the Center for Applied Research in Artificial Intelligence (UD190031RD/25%) grant funded by the Defence Acquisition Program Administration and the Agency for Defence Development.
© 2022 Informa UK Limited, trading as Taylor & Francis Group and The Robotics Society of Japan.
- World robot summit 2020
- home service robot
- modular software framework
ASJC Scopus subject areas
- Control and Systems Engineering
- Human-Computer Interaction
- Hardware and Architecture
- Computer Science Applications