Distributed Control System of Humanoid Robots based on Real-time Ethernet

Abstract

Abstract-In this paper we realize a real-time communication on Ethernet and develop an onbody distributed control system for a humanoid robot, HRP-3P. Real-time communication on Ethernet is realized by (1) a communication method using the data link layer directly and (2) timing control using a real-time operating system ARTLinux. This enables us to reduce the cost of embedded systems and improve developmental efficiency. A CORBA implementation which works on this communication layer is also developed to increase compatibility with existing software. Finally a small-size distributed robot controller is developed for the onbody network of HRP-3P and a distributed I/O system is developed on top of this.

Versatile modular electronics for rapid design and development of humanoid robotic subsystems

“…This reduces the overall system's size and weight as much as possible. To reduce the amount of electrical cables running from the actuators and sensors to a centralised controller, developers have taken a distributed approach [2], [6], [5], where the low level controllers are distributed around the humanoid. This leads to the benefit of reduced cables as well as compactness.…”

Periodic and Aperiodic Communication Techniques for Responsive Link

“…Though the most well-known standard is Ethernet, specialized communication schemes [2,11,5] are required for hard real-time communication. Ethernet is rather effective to soft real-time communication [6].…”

An Embedded Computing Platform for Robot

“…Many researches [4]- [7] indicate that a layered approach is gradually becoming a trend in the design of robotic platform. The benefits of this design method include high-level behavior control, task dispatching and flexible design that can make the control structure of robotic platform more clearly and the operation of robot more efficiently.…”

“…In [4], a layered behavior planning is established for optimizing robot's behavior that helps to modify the behavior model of intelligent robot in accordance with environmental characteristics. In [5]- [7], some group numerous controllers into master/slave control mode and some divide the system into three layers in roughly which include application layer, OS layer and physical layer. These methods are mainly aimed at robotic system and can not only speed up the communication capability of internal system, but also hold the property of reconfiguration and elastic expansion.…”

“…1 is the architecture of embedded computing platform which we propose. Refer to the layered perspective of embedded computing platform in [4]- [7], we describe our robotic embedded computing platform with platform view, system view and robot view respectively.…”