Response of a Neuromechanical Insect
Joint Model to Inhibition of fCO Sensory
Afferents
Clarissa Goldsmith
(B) , Nicholas S. Szczecinski, and Roger D. Quinn
Department of Mechanical and Aerospace Engineering,
Case Western Reserve University, Cleveland, OH 44106, USA
cag111@case.edu
Abstract. This work details the development of a neuromechanical
model of the stick insect femur-tibia joint control network in order to
replicate and explore the “reflex reversal” phenomenon in insect limbs.
We believe that understanding this phenomenon will lead to improved
robotic joint control. We describe the development of this model using
data taken from the extensor half of the network in the insect. To build
a plausible model of the complete joint control network, we additionally mirrored the connectivity of the extensor control networks for the
flexor. We present the results of experiments performed on the network
by selectively inhibiting the system’s sensory afferents in an asymmetrical
manner and observing the behavior of the simulated joint in open and
closed loop scenarios. By inhibiting the network’s flexion position and
velocity afferents, we are able to demonstrate changeover in the joint
from a resistance reflex (RR) to an active reaction (AR) in response to
joint flexion. We discuss why the nervous system might modulate joint
behavior in this manner, as well as how to apply these findings to improve
robotic control.
Keywords: Synthetic nervous system · Drosophibot ·
Neuromechanical model · Reflex reversal
1 Introduction
Insect legged locomotion has been an area of interest in biologically inspired
robotics for many decades. Insects are able to move robustly with relatively
simple nervous systems, providing an approachable template for similar motion
in robots by mimicking their mechanics, neural processes, and behaviors. Furthermore, robots with enough biological fidelity can serve as testbeds for biological hypotheses, improving understanding in both fields. To this end, we
have previously developed Drosophibot, a robot modeled after adult Drosophila
melanogaster including close attention to low level features found in a variety
Supported by the National Science Foundation (Grant Number: 1704436).
c
Springer Nature Switzerland AG 2020
V. Vouloutsi et al. (Eds.): Living Machines 2020, LNAI 12413, pp. 141–152, 2020.
https://doi.org/10.1007/978-3-030-64313-3_15
Joint Model to Inhibition of fCO Sensory
Afferents
Clarissa Goldsmith
(B) , Nicholas S. Szczecinski, and Roger D. Quinn
Department of Mechanical and Aerospace Engineering,
Case Western Reserve University, Cleveland, OH 44106, USA
cag111@case.edu
Abstract. This work details the development of a neuromechanical
model of the stick insect femur-tibia joint control network in order to
replicate and explore the “reflex reversal” phenomenon in insect limbs.
We believe that understanding this phenomenon will lead to improved
robotic joint control. We describe the development of this model using
data taken from the extensor half of the network in the insect. To build
a plausible model of the complete joint control network, we additionally mirrored the connectivity of the extensor control networks for the
flexor. We present the results of experiments performed on the network
by selectively inhibiting the system’s sensory afferents in an asymmetrical
manner and observing the behavior of the simulated joint in open and
closed loop scenarios. By inhibiting the network’s flexion position and
velocity afferents, we are able to demonstrate changeover in the joint
from a resistance reflex (RR) to an active reaction (AR) in response to
joint flexion. We discuss why the nervous system might modulate joint
behavior in this manner, as well as how to apply these findings to improve
robotic control.
Keywords: Synthetic nervous system · Drosophibot ·
Neuromechanical model · Reflex reversal
1 Introduction
Insect legged locomotion has been an area of interest in biologically inspired
robotics for many decades. Insects are able to move robustly with relatively
simple nervous systems, providing an approachable template for similar motion
in robots by mimicking their mechanics, neural processes, and behaviors. Furthermore, robots with enough biological fidelity can serve as testbeds for biological hypotheses, improving understanding in both fields. To this end, we
have previously developed Drosophibot, a robot modeled after adult Drosophila
melanogaster including close attention to low level features found in a variety
Supported by the National Science Foundation (Grant Number: 1704436).
c
Springer Nature Switzerland AG 2020
V. Vouloutsi et al. (Eds.): Living Machines 2020, LNAI 12413, pp. 141–152, 2020.
https://doi.org/10.1007/978-3-030-64313-3_15