Session: 07-11-01 Mobile Robots and Unmanned Ground Vehicles I

Paper Number: 67678

Start Time: Tuesday, 01:10 PM

67678 - Initial Development of Low-Cost Autonomous Rover for Pursuit of Moving Targets 

Autonomous mobile systems have seen steady growth in applications for all physical domains in recent years. One of such domains is ground surfaces that may have rough terrain outside roads built for cars or in harsh environments. The development of unmanned ground vehicles intended for off-road applications received a strong boost during DARPA Grand Challenges about 15 years ago. Since then, a number of advancements were made; and currently, robotic rovers are commercially available, although they can be rather expensive. With the development of simple and cheap electronics platforms, such as Arduino and RPi, as well as a variety of low-cost sensors, a number of projects have been carried out on inexpensive, autonomous and typically small-scale UGVs. The present work belongs to this field. The main ultimate objective is to develop a low-cost autonomous off-road vehicle that can patrol the area and identify, pursue, and effect moving targets. The secondary objective is to create a hands-on educational platform that can be used in robotics and control classes. The current version of UGV under development is based on a four-motor, four-wheeled platform with Arduino-compatible programmable boards and communication modules. This configuration allows users to add a variety of sensors and actuators. This paper presents the hardware/software implementation, including Arduino-compatible electronics architecture, GPS-RTK modules, enhanced compass, telemetry modules, and data logging. The vehicle control modes (manual and autonomous), communication with a moving target, adaptation of software libraries, and recommendations for sensing and steering are described. Two outdoor tests conducted with the autonomous rover are presented in the paper, including trajectory plots and time-dependent kinematic characteristics, such as the relative distance to the target, desired course, and rover heading and speed. In the first trial, the rover successfully reached a set of pre-programmed stationary waypoints. In the second test, the rover autonomously pursued a moving target that had a complicated trajectory. The rover was able to approach and follow the target, while larger deviations of relative distances between the rover and the target and bigger fluctuations of the rover speed were noticed during sudden turns and stops of the target. The future development steps will include implementation of collision avoidance and computer vision modules, so more sophisticated autonomous control methods can be tested and optimized on this platform. The reported in this paper information will be useful for researchers, engineers and educators interested in constructing and using low-cost mobile platforms that can be instrumented with a number of sensors and perform a variety of tasks.

Presenting Author: Cole James Washington State University

Authors:

Cole James Washington State University
Konstantin I. Matveev Washington State University