Home /Research /Atlas Humanoid Robot Control with Flexible Finite State Machines for Playing Soccer
MANIPULATION

Atlas Humanoid Robot Control with Flexible Finite State Machines for Playing Soccer

Davis S. Catherman, Jakub Tomasz Kaminski, Aishwary Jagetia

Year
2020
Citations
2

Abstract

This work concerns Atlas robot autonomously com-pleting core soccer tasks in a static environment. A soccer ball is localized using robot 2D and 3D perception. The target kicking position is reached in a collision-free manner using robot locomotion and navigation capabilities. The ball is kicked thanks to leg manipulation motion planning. Additionally, FlexBE is used to implement a user in the loop state machine. Generic custom open-source states are used throughout the project and their functionality is tested. An overview of subject-relevant literature, techniques used and experiments performed are provided. We also discuss the system and environment-based limitations in the core project areas.

Keywords

RobotHumanoid robotComputer scienceFinite-state machineAtlas (anatomy)Ball (mathematics)SimulationArtificial intelligenceRobot controlHuman–computer interaction

Related papers

Browse all MANIPULATION papers