Session: 08-11-01: Mobile Robots and Unmanned Ground Vehicles I

Paper Number: 165390

From RC Vehicles to Collaborative Robotics: A Low-Cost Approach to UGV Development 

The advancement of collaborative robotic systems necessitates the development of cost-effective, modular platforms capable of supporting a diverse range of payloads. This work details the development of the Crover, which is designed to perform in real-world environments where individual units can support varying payloads, maintain computational efficiency, and remain cost-effective to enable the development of multiple systems for researching application of collaborative multiple-robot systems (MRS). Ensuring compatibility with open-source Linux-based software environments to enable broader accessibility and customization was critical to ensure system flexibility when integrating with sensors and payloads. Existing commercially available robotic platforms were found to be either cost-prohibitive, mechanically inadequate, or lacked necessary software flexibility. A review of prior research suggested that commercial off-the-shelf (COTS) remote-controlled vehicle platforms could provide a viable alternative for developing low-cost, high-performance robotic systems. Based on this insight, a remote-controlled rock crawler chassis was selected as the foundation for the Crover due to its mechanical robustness and affordability. The chassis underwent extensive modifications to enable closed-loop control feedback, the integration of a power distribution and management system, and structural support for various payloads and computational subsystems. Additionally, this work discusses the systematic modifications and enhancements made to the base vehicle chassis to instrument precise kinematic control. The deployment of a distributed control architecture supported modular electro-mechanical payload integration, allowing for flexible system customization. The software design leverages a distributed control framework to accommodate additional computational resources, such as artificial intelligence accelerators, vision systems, and microcontrollers, thereby extending the functional capabilities of the robotic system. A series of experiments were conducted to develop, validate, and refine the kinematic controller of the Crover. First of all, the initial kinematic modeling was performed through simulation, and empirical testing was subsequently conducted to fine-tune and validate the control system’s performance. These tests provided critical insights into the efficacy of the closed-loop control system and the overall mechanical and electrical reliability of the platform. Secondly, testing of Crover was conducted in a laboratory setting to confirm full integration and utilization of the kinematic controller and the electro-mechanical integration of its intended research payload. Lastly, testing was conducted in outdoor environments to further validate the Crover performance in a real-world environment. The findings of this study indicate that the Crover framework serves as a viable approach for developing cost-effective, modular, and adaptable unmanned ground vehicles (UGVs) using commercially available vehicle chassis and open-source software. Furthermore, the methodologies employed in this research can be extended beyond ground-based applications to the development of unmanned aerial vehicles (UAVs) and unmanned surface vehicles (USVs). This work underscores the potential of leveraging COTS components and open-source software in the development of autonomous robotic platforms, significantly reducing research costs while expanding the possibilities for studying collaborative robotic systems in diverse operational environments.

Presenting Author: Bryan Cochran Georgia Institute of Technology

Presenting Author Biography: Bryan Cochran recently obtained his Ph.D. in Mechanical Engineering with a primary focus in Robotics and Automation from Georgia Institute of Technology. His research focuses on the design and development of robotic systems and exploration of novel communication systems designed for deployment in collaborative multiple robot systems. Bryan leverages recent developments of commercially available hardware coupled with open-source software to expedite the development and deployment of adaptable robotic platforms.

Authors:

Bryan Cochran Georgia Institute of Technology
José Levy Georgia Institute of Technology
Viraj Pahwa Georgia Institute of Technology
Bert Bras Georgia Institute of Technology