日本語

 

Update(MM/DD/YYYY):11/21/2003

Knowledge Distributed Robot Control System Using IC Tags

- Information necessary for robot manipulation is distributed throughout the human everyday environment -

Key Points

  • The current robot control based on artificial intelligence and image processing technology has reached the limit for ensuring robot working in the everyday environment where the presence of diverse objects makes it extremely difficult to recognize the real “world” and to generate the work sequences.
  • It is proposed to allocate data required for a robot to work in the objects-strewn ambience in a distributed manner by putting an IC tag to each of objects.
  • To allocate the information for robot manipulation in a distributed manner, we propose to fix an IC tag to each object in the object-scattered environment.
  • Each IC tag has the network address of its maker, which makes it possible for a robot to acquire the object information through the network.
  • The proposed scheme provides the object information to the conventional image processing technology. Thanks to this, the conventional vision system can achieve more accurate object localization and much easier map generation in our everyday environment.
  • Distributed knowledge system not only can effectively reduce robot makers’ programming burden, but also provide diversification to the robot.


Synopsis

In the present robot industry, not only robots for industrial purpose but also those for serving in the human living environment are being eagerly sought for. The current robot control based on artificial intelligence and image processing technologies cannot support the robot works in the actual environment, because of difficulty in fully recognizing diverse objects placed in the living environment.

The Intelligent Systems Research Institute (ISI) of the National Institute of Advanced Industrial Science and Technology (AIST), one of independent administrative institutions, has proposed a distributed knowledge robot control scheme in which every object is posted with an IC tag containing the manufacturer’s network address and the knowledge information required for a robot to handle the object concerned (patent filed). The scheme allows the robot to recognize objects more easily and the manufacturers to have burden of robotic programming reduced. The ISI has successfully developed a demonstration plant by using available IC tag, image processing and robot control technologies.

The developed plant concerned a robot’s work to prioritize (or decide the order of priority) table wares randomly placed on a table. The objects scattered all over the table are automatically tidied out thanks to the IC tag readers installed on the table, while generating a robot operating strategy based on the obtained information. The system successfully demonstrated the effectiveness of the proposed scheme.

It is expected that the new system will significantly reduce the burden of programming on the part of robot systems manufacturers, and largely accelerate the development of robots working under the human living environment.

It is intended to extend the application of this control technique to general-natured works to verify further its competence.

The proposed system will be disclosed to academic and industrial partners interested in making collaboration with the ISI/ AIST, at the OPEN HOUSE 2003, an outcome exhibition of the Institute, to be held from 09:30 to 16:00 on Thursday, November 13, 2003, at the AIST Tsukuba Center.


Fig1.
A concept diagram of “Distributed Knowledge Type Robot Control”





▲ ページトップへ