Hardware and software architecture for intelligent robot control-an approach to dual control architecture

A new “dual control” robot control architecture is proposed. This control architecture endows a robot with human-like intelligent control capability similar to that of the human cerebrum. This improves robot adaptability to unpredictable circumstances, and allows a higher level of adaptability than conventional sensory feedback control, which depends on robot local sensory information, and which is limited to functions similar to those of the human cerebellum. With dual control, robot sensor data are processed qualitatively to recognize and understand global environmental situations, and decisions for real-time robot actions are then made. A closely coupled dual processor system with a dual bus system is proposed as the control architecture. The software system implemented for the hardware, as well as an example of the control, are also described