It is important for a robot to appropriately respond to its surrounding environment and events to communicate smoothly with humans as opposed to following fixed movements specified in advance. In this paper, a simple bi-layered architecture (SB architecture) is proposed to integrate behaviors in two stages: prioritization and weighted averaging. In addition, SB architecture integrates voluntary movements by prioritization and integrates involuntary and reflex movements by weighted averaging, thereby generating robot behaviors that immediately respond to events that occur around the robot. What advantages SB architecture has are that it can easily generate various behaviors by combining multiple behaviors and that it is possible to propose a simple design for robot behavior. Furthermore, since behavior occurs in response to sensors, the robot's behaviors are reactive to the surrounding environment and events. In particular, if a robot performs the behavior necessary for communication with humans, such as changing its gaze and gestures, it is possible to ensure that the robot possesses liveness to promote communication with humans. In addition, since SB architecture has a simple structure to facilitate the design of robots, it is possible to automatically set parameters by optimizing the parameters using pairs of sensor information and ideal behaviors. The optimization of the parameters leads to generating behaviors characterized with liveness by appropriately combining behaviors.
|Title of host publication
|IROS 2017 - IEEE/RSJ International Conference on Intelligent Robots and Systems
|Institute of Electrical and Electronics Engineers Inc.
|Number of pages
|Published - 2017 Dec 13
|2017 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2017 - Vancouver, Canada
Duration: 2017 Sept 24 → 2017 Sept 28
|2017 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2017
|17/9/24 → 17/9/28
ASJC Scopus subject areas
- Control and Systems Engineering
- Computer Vision and Pattern Recognition
- Computer Science Applications