The Planetary Rover project is developing the Ambler, a novel legged robot, and an autonomous software system for walking the Ambler over rough terrain. As part of the project, we have developed a system that integrates perception, planning, and real-time control to navigate a single leg of the robot through complex obstacle courses. The system is integrated using the Task Control Architecture (TCA), a general-purpose set of utilities for building and controlling distributed mobile robot systems. The walking system, as originally implemented, utilized a sequential sense-plan-act control cycle. This report describes efforts to improve the performance of the system by concurrently planning and executing steps. Concurrency was achieved by modifying the existing sequential system to utilize TCA features such as resource management, monitors, temporal constraints, and hierarchical task trees. Performance was increased in excess of 30 percent with only a relatively modest effort to convert and test the system. The results lend support to the utility of using TCA to develop complex mobile robot systems. (author)
Concurrent Planning and Execution for a Walking Robot
1990
19 pages
Report
No indication
English
Robotics/Robots , Computers, Control & Information Theory , Control , Planning , Robots , Architecture , Computer programs , Distribution , Hierarchies , Mobile , Monitors , Perception , Real time , Resource management , Roughness , Self operation , Sequences , Terrain , Trees , Walking , Robot dynamics
Concurrent planning and execution for a walking robot
IEEE | 1991
|Walking robot, method for controlling walking robot, and walking robot system
European Patent Office | 2023
|Trajectory Planning for Multi-arm Space Walking Robot
AIAA | 2015
|