Developing Adjustable Walking Patterns for Natural Walking in Humanoid Robots

Because of the potential for interaction with humans, research in humanoid robotics has made significant progress in recent years. It has long been imagined that some day, humanoid robots will serve people with disabilities, improve search-and-rescue efforts in dangerous conditions, and perhaps even replace humans in doing our everyday tasks. The ability of a humanoid robot to do these tasks, however, relies on its locomotion capabilities. On a large scale, the robot must be able to walk in many types of terrain if it is to operate successfully in diverse conditions. A walking pattern was thus developed that allows for the stride length, the single-support balance point, turning angle, and the maximum height of the swing foot to be modified dynamically. This will then enable the robot to undergo a learning process to determine how certain parameters affect its walking and will enable higher-level control systems that use inputs from vision and tactile sensors to make its walking robust to many different environments.