ROAMAN is a computer program for autonomous navigation of a mobile robot on a long (as much as hundreds of meters) traversal of terrain. Developed for use aboard a robotic vehicle (rover) exploring the surface of a remote planet, ROAMAN could also be adapted to similar use on terrestrial mobile robots. ROAMAN implements a combination of algorithms for (1) long-range path planning based on images acquired by mast-mounted, wide-baseline stereoscopic cameras, and (2) local path planning based on images acquired by body-mounted, narrow-baseline stereoscopic cameras. The long-range path-planning algorithm autonomously generates a series of waypoints that are passed to the local pathplanning algorithm, which plans obstacleavoiding legs between the waypoints. Both the long- and short-range algorithms use an occupancy-grid representation in computations to detect obstacles and plan paths. Maps that are maintained by the long- and short-range portions of the software are not shared because substantial localization errors can accumulate during any long traverse. ROAMAN is not guaranteed to generate an optimal shortest path, but does maintain the safety of the rover.
This program was written by Terrance Huntsberger and Hrand Aghazarian of Caltech for NASA's Jet Propulsion Laboratory. For further information, access the Technical Support Package (TSP) free online at www.techbriefs.com/tsp under the Software category.
This software is available for commercial licensing. Please contact Karina Edmonds of the California Institute of Technology at (818) 393-2827. Refer to NPO-30532.
This Brief includes a Technical Support Package (TSP).
Autonomous Navigation by a Mobile Robot
(reference NPO-30532) is currently available for download from the TSP library.
Don't have an account?
Overview
The document is a Technical Support Package from NASA’s Jet Propulsion Laboratory (JPL) detailing advancements in autonomous navigation for mobile robots, specifically focusing on the Field Integrated, Design and Operations (FIDO) rover. This rover serves as a terrestrial analog for the Mars Exploration Rovers (MER) set to launch in 2003. The primary goal of the research is to enhance rover autonomy for planetary surface operations, enabling the rover to navigate and perform tasks with minimal ground control interaction.
The FIDO rover is equipped with a sophisticated suite of sensors, including an inertial navigation unit (INU) with 3D gyros and accelerometers, a sun sensor, and various imaging systems. These sensors facilitate multisensor fusion through an Extended Kalman Filter (EKF) approach, which, combined with pattern recognition and tracking algorithms, allows for improved navigation and planning capabilities. The document discusses the challenges of rover navigation, particularly for long-duration missions to Mars, where the rover must autonomously navigate to destinations that may be hundreds of meters away, perform precision rendezvous with targets, and deploy scientific instruments without human intervention.
The paper outlines a three-phase navigation strategy for the rover, which includes long-range, mid-range, and short-range approaches to target acquisition. For long-range navigation, the rover utilizes line features and wavelet-derived texture signatures to identify man-made structures, such as landers, from significant distances. As the rover approaches its target, it refines its navigation using more detailed features to ensure precise alignment for tasks like sample collection.
The document also highlights the results of validation studies conducted in various remote environments, such as the Mojave Desert and Central Nevada, where the algorithms were tested to ensure their effectiveness in real-world scenarios. The FIDO rover's design incorporates a six-wheel rocker-bogie suspension system, allowing it to traverse obstacles up to 30 cm in height, making it suitable for the rugged terrain of Mars.
Overall, this Technical Support Package emphasizes the importance of developing autonomous navigation technologies for future space exploration missions, showcasing JPL's commitment to advancing robotics and enhancing the scientific return from planetary exploration.
