Trunk Pitch Oscillations for Joint Load Redistribution in Humans and Humanoid Robots

Abstract

Creating natural-looking running gaits for humanoid robots is a complex task due to the underactuated degree of freedom in the trunk, which makes the motion planning and control difficult. The research on trunk movements in human locomotion is insufficient, and no formalism is known to transfer human motion patterns onto robots. Related work mostly focuses on the lower extremities, and simplifies the problem by stabilizing the trunk at a fixed angle. In contrast, humans display significant trunk motions that follow the natural dynamics of the gait. In this work, we use a springloaded inverted pendulum model with a trunk (TSLIP) together with a virtual point (VP) target to create trunk oscillations and investigate the impact of these movements. We analyze how the VP location and forward speed determine the direction and magnitude of the trunk oscillations. We show that positioning the VP below the center of mass (CoM) can explain the forward trunk pitching observed in human running. The VP below the CoM leads to a synergistic work between the hip and leg, reducing the leg loading. However, it comes at the cost of increased peak hip torque. Our results provide insights for leveraging the trunk motion to redistribute joint loads and potentially improve the energy efficiency in humanoid robots.

Trunk pitch oscillations for energy trade-offs in bipedal running birds and robots

“…A VP A in the human TSLIP model on the other hand, leads to the backward (anteroposterior) trunk motion during the stance phase [23], which is not consistent with the oscillation direction observed in human running [24]. It is shown in [8] that placing the VP below the CoM (VP B ) generates the forward (posteroanterior) trunk motion that is consistent with the experiments. In this paper, we inquire whether a similar relation exists for the avian morphology.…”

“…In order to simplify and guide the parameter search, we use the iterative gait generation framework in [8]. The framework has an initial controller that combines the VP based torque in Equation (2) with a PID controller on the pitch angle, which yields stable gaits with semifocused GRF in Figure 5c The TSLIP model has seven morphological parameters, which are selected to match an ostrich of 80 kg with 1 m leg length (see Table 1), so that the avian model parameters are closer to the human model in [8] to enable comparison. Figure 2: A basic TSLIP model leads to non-zero leg length velocity at touch-down ( ), which is not consistent with the measurements in the avian gait ( ).…”

“…The non-continuities in leg damping are not realistic for the touch-down, however, are common for take-off in avian gaits due to the early leg take-off. Consequently, we use the bilinear leg damper in [8] that combines the leg length velocity with leg deflection ( ) to obtain non-zero damping forces at touch-down. Figure 3: The basic leg take-off condition for the TSLIP model (l ≥l 0 ) can cause the CoM to reach its apex during the stance phase avian TSLIP model.…”

“…the VP and trunk, and functions for all situations. Figure 5: Simulation setup and resulting GRF patterns for avian TSLIP model, which is generated with the control framework presented in [8]. The estimated VP is marked with black rimmed circle.…”

“…One of our aims was to identify the consequences of having a pronograde trunk in comparison to an orthograde one. For this reason, we recreate the data in [8] for a human TSLIP model. We select the model parameters to match a human with a mass of 80 kg, leg length of 1 m and a trunk angle of 10 • .…”