Comparison of trajectory generation methods for a human-robot interface based on motion tracking in the Int²Bot

Abstract

The acceptance of artificial devices like prostheses or other wearable robots requires their integration into the body schemas of the users. Different factors induce, influence and support the integration and acceptance of the device that substitutes or augments a part of the body. Previous studies have shown that the inducing and maintaining factors are visual, tactile and proprioceptive informations as well as their multisensory integration. This paper describes the vision-based part of the human-robot interface in the Int 2 Bot, which is a robot for the investigation of lower limb body schema integration during postural movements. The psychological approach and the technical setup of the robot, which is designed to imitate postural movements in the sagittal plane to imitate the human subject while performing squats, are outlined. To realize the imitation, an RGB-D sensor, in form of a Microsoft Kinect, is used to capture the subjects motions without contact and thereby avoid disturbances of body schema integration. For generation of the desired joint trajectories to be tracked by the control algorithm, different methods like an extended Kalman filter, inverse kinematics, an inverse kinematics algorithm using Jacobian transpose and approaches based on kinematic assumptions are presented, evaluated and compared based on human data. Benchmarking the results with data acquired using a professional motion capturing system shows that best overall joint angle estimations are achieved with the extended Kalman filter. Finally, the practical implementation within the robot is presented and the tracking behavior using the trajectories generated with the extended Kalman filter are analyzed.

Implementation, control and user-feedback of the Int²Bot for the investigation of lower limb body schema integration

“…Although limited by knee actuation, the robot can perform considerably fast squats with 0.8 Hz, which comes up to maximum human capabilities. Evaluating the robot with a human user shows that the setup is currently limited by trajectory generation due to delays caused by errors in RGB-D skeleton tracking and the related data transfers [21]. Hence, only preliminary body schema integration experiments can be performed.…”

“…To control the robot, an RGB-D sensor performs contactless human motion tracking and the desired robot trajectories are generated from this. This vision-based part of the human-robot interface is extensively dealt with in [21]. Control and compensation algorithms further utilize proprioceptive data from the robot joints.…”

“…Additional sensing of human data is provided by an incremental encoder measuring knee position and velocity. Those can be included in trajectory generation algorithms to improve results obtained using the RGB-D sensor only [21]. An intuitive haptic feedback might be reached by vibrations scaled proportionally to the mechanical knee load in amplitude or frequency.…”

“…As these significant delays between human and robot motion exceed the requirements from [13], only preliminary body schema integration experiments can be performed so far. To reduce the delays, measurements of the knee orthosis can be fused with RGB-D data or improved motion tracking can be introduced, as the delays result from high estimation errors in skeleton tracking [21].…”

Estimación de la cinemática de las articulaciones de miembro inferior por medio del filtro de Kalman extendido

“…Cabe resaltar que los diversos procedimientos propuestos buscan desarrollar y evaluar un método que permita la identificación rápida y fiable de los parámetros inerciales de los segmentos corporales [18]. Así se destaca la aplicación de diversos métodos como se menciona en [19], [20] y [21], entre los cuales sobresale el uso de sistemas basados en fusión de acelerómetros y giróscopos [22], [23], técnicas de fotogrametría [1], [24], y aplicación de estereofotogramas y placas de fuerza como medios para lograr la determinación en tiempo real de los ángulos de las articulaciones, las velocidades, las aceleraciones y el torque de las articulaciones de un ser humano, además de las longitudes de segmentos específicos y parámetros inerciales.…”