Abstract
In generally, many robot designers decide a walking algorithm separately from how to sense the around environment. Therefore it is difficult to calculate the gait plan when the estimated environment included some errors. Our study is originated from an idea that we have to integrate the model design and the sensing strategy into a walking robot model. When the robot is walking on the ground, the robot is affected from environment. If environment is different, the robot is affected differently. If the robot is affected differently, the gait of the robot shifts. This paper describes the concept that a walking robot is able to sense a shift of the ground through its walking conditions. Some simulation results show the effectiveness.
