Gait generation based on Mimic-Reactive Architecture

A new architecture called the “Mimic-Reactive Architecture” for the design of autonomous robots is presented. It is a practical methodology aimed at the easy design of autonomous robot systems in order to achieve enhanced survival. Experiments with a real legged robot are presented in order to probe their validity as methodologies for agent design. In the proposed method, the higher-level agents generate “mimic information” to use the reactive rules of the lower-level agents. The information is applied to sensory input. Mimic-reactive systems act according to such mimic information. As a result, the system avoids failure through critical impacts caused by imprudent implementation of higher-level agents, because the reactive engine in the lower-level agents gives sufficient survivability and reactivity and there is no direct connection between the higher-level agents, which is usually complex and needs a certain amount of time to process