Real-Time Navigation for Bipedal Robots in Dynamic Environments

Loading...
Thumbnail Image

Date

2022-12

Journal Title

Journal ISSN

Volume Title

Publisher

The Ohio State University

Research Projects

Organizational Units

Journal Issue

Abstract

The popularity of mobile robots has been steadily growing, with these robots being increasingly utilized to execute tasks previously completed by human workers. Bipedal robots, a subset of mobile robots, have been a popular field of research due to the large range of tasks for which they can be utilized. For bipedal robots to see a similarly successful integration into society, robust autonomous navigation systems need to be designed. These autonomous navigation systems can generally be divided into three components: perception, planning, and control. A holistic navigation system for bipedal robots must successfully integrate all three components of the autonomous navigation problem to enable robust real-world navigation. Many works expand on fundamental planning algorithms such as A*, RRT, and PRM to address the unique problems of bipedal motion planning. However, many of these works lack several components required for autonomous navigation systems such as real-time perception, mapping, and localization processes. Thus, the goal of this research is to develop a real-time navigation system for bipedal robots in dynamic environments which addresses all components of the navigation problem. To achieve this: a depth-based sensor suite was used for obstacle segmentation, mapping, and localization. Additionally, a two-stage planning system generates collision-free and kinematically feasible trajectories robust to unknown and dynamic environments. Finally, the Digit bipedal robot's default low-level controller is used to execute these feasible trajectories. The navigation system was first validated on a differential drive robot in simulation. Next, the system was adapted for bipedal robots and validated in hardware on the Digit bipedal robot. In both simulation and in hardware experiments, the implemented navigation system facilitated successful navigation in unknown environments and in environments with both static and dynamic obstacles.

Description

Keywords

Motion Planning, Humanoid Robots, Autonomous Navigation, Bipedal Robots, Real-Time Planning

Citation