International Journal of Humanoid Robotics

Robotic feet play a fundamental role in the walking performance of a biped robot. Feet are essential to maintain dynamic stability and to propel the body during walking. They may ensure stability on uneven terrains. Yet, complex feet are seldom used on humanoids. This paper surveys 36 types of robotic feet we found in the literature. We classified them according to strategy, capabilities, structure, number of degrees of freedom, actuation method of ankle and foot, type of actuator, sensorization and type of control. Subsequently, we analyzed the dynamic and static models of flexible feet. We discussed considerations on foot dynamics or kinematics in the robot’s whole body control system. We analyzed both active joints control for feet including actuated joints, and control for feet with elastic elements (for example, a rubber layer in the sole). Finally, we present some limitations of robotic feet and possible future developments.

Legged robots have been interesting for researchers because they can move on uneven terrain and operate in dangerous areas for humans. Among these robots, the single-legged robot is an excellent choice to study because they are simpler than another legged robot. In this paper, we have designed a single-legged robot similar to the anatomy of the human leg. It consists of four active joints and one passive joint. The hip, knee, and ankle joint are considered active joints, and the toe joint is a passive joint. To make the simulation more accurate, we should define the physical property of the environment in detail. So, ground contact and friction model should be specified. After defining the robot model, we designed joint motion. Path planning aims to make the robot jump in the vertical direction. Besides, the robot must start jumping on the toe joint. The toe joint stiffness and damping play the leading role in jumping height. So, there should be optimum values for these parameters. The genetic algorithm is then used to optimize jumping height.

Most artificial muscles currently in existence driven by negative pressure are in the form of origami. Their contractions are directly attained by extracting the air internal, and obtaining a differential pressure with the external (0.101MPa). During contraction, the artificial muscle simultaneously encounters radial and axial pressures provided by the atmosphere. If the muscle is isotropic, the radial pressure would make the artificial muscle unreliable, which would weaken the muscle’s contraction, including contraction force, ratio, and speed. This paper first introduces the contraction theory and dysfunction of the muscle. Then, an optimization method is proposed as local reinforcement, aiming at the muscle’s dysfunction that may occur during contraction. Then, by analyzing and comparing the optimization effects on the muscle contraction force and deformation with different reinforced regions, the endplates and crests are determined to be appropriate. Subsequently, the deformations of the reinforced regions when stressed are analyzed, and the adequate reinforcement dimensions are confirmed. Finally, the muscle’s model passes the simulation for ensuring its contraction without dysfunction, demonstrating the rationality of the reinforcement method proposed in this paper. Hereby, the goal of reinforcing the muscle’s radial stiffness is achieved, making the artificial muscle generate an optimal contraction performance.