| Research Abstract |
In the conventional design of human-machine systems, the designer of the system has taken an external position lying outside of the interactions made between the human user and the automated system. He has been an external observer and his task has been on how to design the optimized inter acti on between them based on the technology-centered idea. Anyway, human and computer should be looked as two parties coming from the same positions and having different expertise. The human and the automation system must be regarded as equivalent partners, rather than one of them is regarded as superior to the other. It may be still usual that there happens a contradictory situation where the designer and the user of the automation system meet conflicts, but this must be allowed. What is more important here is how to let them cooper ate with each other to change a current conflictual status into a new coordinated one that can lead to a safer plant status and the better operations. In this sense, we
… More
have to extend our views on a human-computer relationship toward a socially-interacting relationship between the human and the machine to establish a consensual domain between the machine and the human. By interacting with the others and reflecting on this experience, living systems reconstruct a definition of the reality, which causes them behave differently and changes the interaction patterns. This is iterated indefinitely, so the living system keeps changing and produce a variability.
Based on the above idea, in this project we developed a general learning model of observing systems, rather than observed systems, and have a continuous self-production process enabled by evolutional development through its internally-inspired activities. The actual domains attempted in this project are as follows :
1. Interface Agent as a Human Associate
2. Robot Teleoperation System Using Task-Morphology
3. Mixed-Initiative Interactions for Robot-Teleoperation
4. Selective and Dynamic Perception of Interface Agents
5. Interface Agent's Managing Computational Complexity through Activating Multigranular Categories for Concept Sharing
6. Ecological approach to analysis and design of Human-Machine Interactions
7. Emergence of Organizational Behaviors out of Group Works under Time Pressure and Resource Boundedness Less
|
