This paper proposes a new rehabilitation robot for human gait training. As humans change walking speed, their nervous systems adapt muscle activation patterns to modify arm swing for the appropriate frequency. The faster the speed, the higher frequency resulted, while the lower the walking speed, the lower the arm swing rate frequency is. By analyzing both properties, we can find the relation between arm swing and lower limb motion. This rehabilitation robot is being developed by applying the concept that the lower limb motion is generated by the arm swing. The lower limb part consists of a slider mechanism and a 3 DOF footpad. This mechanism allows walking in a virtual environment with uneven and various terrains, so that can improve the versatility of rehabilitation training. The upper limb consists of arm swing handle as user input. The analog handgrip on the each side of the handle is provided for the patients who cannot swing both of the arms due to illnesses or weaknesses. For haptic sense, the force sensor is located on the top of the footpad to measure the human foot intention, and the feedback motor is located between the arm handle. At the final stage of development, virtual reality environment is proposed.
