Reflex Control of Robotic Gait Using Human Walking Data

Abstract

Control of human walking is not thoroughly understood, which has implications in developing suitable strategies for the retraining of a functional gait following neurological injuries such as spinal cord injury (SCI). Bipedal robots allow us to investigate simple elements of the complex nervous system to quantify their contribution to motor control. RunBot is a bipedal robot which operates through reflexes without using central pattern generators or trajectory planning algorithms. Ground contact information from the feet is used to activate motors in the legs, generating a gait cycle visually similar to that of humans. Rather than developing a more complicated biologically realistic neural system to control the robot's stepping, we have instead further simplified our model by measuring the correlation between heel contact and leg muscle activity (EMG) in human subjects during walking and from this data created filter functions transferring the sensory data into motor actions. Adaptive filtering was used to identify the unknown transfer functions which translate the contact information into muscle activation signals. Our results show a causal relationship between ground contact information from the heel and EMG, which allows us to create a minimal, linear, analogue control system for controlling walking. The derived transfer functions were applied to RunBot II as a proof of concept. The gait cycle produced was stable and controlled, which is a positive indication that the transfer functions have potential for use in the control of assistive devices for the retraining of an efficient and effective gait with potential applications in SCI rehabilitation.

Toward a dynamic analysis of bipedal robots inspired by human leg muscles

“…Biarticular muscles: these muscles are connected to two links separated by a third link and two joints and can drive these two joints simultaneously, and have the capability to transfer energy mainly produced by monarticular muscles to joints where they can effectively contribute to the desired aim of movement. The mono and biarticular muscles that correspond to rectus femoris (RF), biceps femoris (BF) gastrocnemius (GAS) and tibialis anterior (TA) in human legs play significant roles in the stability of human walking [10,11].The muscles (RF) is a bifunctional muscle responsible for hip flexion in the swing phase and knee extension in the late swing and stance phase, the (BF) muscle responds to the hip extension in the stance phase and the knee flexion in the swing phase the (GAS) muscle is primarily responsible for ankle plantarflexion but also takes a minor role in knee extension, the TA muscle has two distinct roles during human walking: (1) to dorsiflex the ankle during the swing phase for foot clearance and placement; (2) to contract during ankle plantarflexion at the initial foot contact with the ground [1,15]. Thus, using springs like mono-biarticular muscles in the leg mechanism of bipedal robots are able to achieve the stability of human-like leg walking compared with those that assume the bipedal robots like rigid body structures.…”

Bipedal robotic walking control derived from analysis of human locomotion

“…2 , the muscle activity can be identified as following the ipsilateral HS for the hip extension and ipsilateral TO for the knee flexion (Macleod et al. 2014 ).…”

“…In the previous study the muscle transfer functions were optimised using a curve fitting process to remove spurious artefacts and resampled at a specific sampling frequency to fit the mechanical system of the RunBot II (Macleod et al. 2014 ). Conversely, in this paper, a second-order low-pass Bessel filter; see Eq.…”

“…The calculation of transfer functions relating sensory information and muscle EMG has been discussed in detail in our previous study (Macleod et al. 2014 ) where the transfer functions were applied to a prior generation of the RunBot (RunBot II) as a proof of concept. The stable gait cycle generated indicated that the approach had potential for use in robotic control.…”

Adaptive multichannel FES neuroprosthesis with learning control and automatic gait assessment

“…Meng et al extended this approach to include four muscles, the quadriceps, hamstrings, tibialis anterior and gastrocnemius muscles [16]. In a previous study [17], the EMG muscle activity of ten healthy subjects during gait was recorded in relation to five gait events. This recorded EMG activity was converted to a stimulation intensity pattern and played back at the corresponding gait events of each individual participant.…”