
Credit: Jimmy Woo.
Researchers from Stanford University, Princeton University, and Dexterity recently developed TidyBot++, a holonomic mobile robot that can perform a variety of household tasks and can be useful for training and testing new algorithms for robotics applications.
This promising robotic platform was presented at the Robot Learning Conference and outlined in a paper published on the arXiv preprint server.
“Our paper was inspired by the growing need for scalable data collection in the field of robotics, particularly mobile operations,” Janet Borg, supervising author of the paper, told Tech Xplore.
“While imitation learning has shown promising results, its success is highly dependent on the availability of real-world demonstrations, which are difficult to obtain due to the lack of suitable and accessible research platforms. We noticed a gap in the availability of open source, flexible and cost-effective hardware for mobile operation in home environments, which is the main focus of many robotics applications. It’s clothing.”
The main aim of recent research by Borg et al. is to contribute to advances in robotics research by developing a new robotic platform that is suitable for tackling household chores and can also be used to collect training data for imitation learning algorithms. That was the case. . To foster wide adoption within the robotics community, this platform was designed to be easily accessible, robust, and versatile.
“TidyBot++ is a mobile robot designed to perform household tasks accurately and easily,” explained Jimmy Wu, lead author of the paper.
“At its heart is a holonomic base with motorized casters that can move freely in all directions, including forwards and backwards, sideways and rotationally, without the constraints found in traditional robot bases such as differential drive systems. “The robot’s movements are simplified, making tasks such as repositioning in tight spaces much more efficient.” ”
The TidyBot++ robot has three main components. The first is a modular frame made from aluminum T-slot extrusions that is easy to build and can be customized by adding different robotic arms and sensors.
The aluminum frame incorporates electric caster wheels with a small “caster offset”. These wheels allow you to control all planar degrees of freedom simultaneously, allowing you to rotate the robot essentially like a motorized office chair.
The third component of the robot consists of a mobile phone telecontrol interface that leverages the WebXR API for real-time pose tracking. Using this interface, users can intuitively control the robot’s movements just by moving their smartphone.
“Compared to other platforms, TidyBot++ is easy to navigate, accessible, and optimized for home learning,” Borg said. “The holonomic base ensures efficiency and smooth operation even in cluttered environments.”

Credit: Jimmy Woo.

Credit: Jimmy Woo.
Compared to other advanced robotic platforms, TidyBot++ is relatively low-cost, with manufacturing costs ranging from $5,000 to $6,000. Its main structure can be easily customized by adding specific robotic arms and sensors, making it easier to utilize for research.
To test the robot’s potential, Borg and his colleagues have already conducted a variety of real-world experiments. Their findings are very promising, as by executing the trained policies in several demonstrations, TidyBot++ was able to perform a variety of household tasks, such as opening the refrigerator, wiping down the countertop, and loading the dishwasher. It has been shown that it can be completed with a high success rate.
“We found that the holonomic base significantly outperformed non-holonomic designs in terms of efficiency, task success rate, and ease of remote control,” Wu said. “By making the design and software completely open source, we aim to democratize access to our robust mobile manipulators and enable researchers around the world to advance their mobile manipulator research.”
In the future, the robots developed by this team of researchers could help collect high-quality data to train AI algorithms for robotics through imitative learning. TidyBot++ can also be deployed in educational settings as a platform to test these algorithms, and ultimately in home environments to assist users with household chores.
“In our next research, we plan to extend the capabilities of TidyBot++, for example by integrating advanced sensors and multiple arms to expand the range of household tasks TidyBot++ can perform,” Borg said. “We also plan to develop more intuitive interfaces and automation capabilities for collecting large datasets to learn complex operational policies.”

Credit: Jimmy Woo.

Credit: Jimmy Woo.
To facilitate the widespread adoption of their robots, the researchers are currently trying to improve their designs by addressing some of the limitations, such as backward maneuverability (i.e., currently it is not possible to use external forces to move the robots). is difficult). We also want to conduct tests to investigate whether robots can easily collaborate with other robots to tackle missions that can be accomplished more effectively as a team.
“In the future, we will also investigate ways to train policies that generalize across different environments and tasks with minimal additional training,” Wu added.
“By continuing to improve TidyBot++, we aim to push the boundaries of what mobile manipulators can achieve and contribute to making robotic assistance more accessible and impactful in everyday life.”
Details: Jimmy Wu et al, TidyBot++: An Open-Source Holonomic Mobile Manipulator for Robot Learning, arXiv (2024). DOI: 10.48550/arxiv.2412.10447
Magazine information: arXiv
© 2025 Science X Network
Citation: Open-source holonomic mobile manipulator could advance home robotics (January 9, 2025) https://techxplore.com/news/2025-01-source-holonomic-mobile-advance- Retrieved January 9, 2025 from household.html
This document is subject to copyright. No part may be reproduced without written permission, except in fair dealing for personal study or research purposes. Content is provided for informational purposes only.