Simulation of the Arthropod Central Complex: Moving Towards Bioinspired Robotic Navigation Control

Abstract

It is imperative that an animal have the ability to track its own motion within its immediate surroundings. It gives the necessary basis for decision making that leads to appropriate behavioral responses. It is our goal to implement insect-like body tracking capabilities into a robotic controller and have this serve as the first step toward adaptive robotic behavior. In an attempt to tackle the first step of body tracking without GPS or other external information, we have turned to arthropod neurophysiology as inspiration. The insect brain structure called the central complex (CX) is thought to be vital for sensory integration and body position tracking. The mechanisms behind sensory integration are immensely complex, but it was found to be done with an elegant neuronal architecture. Based on this architecture, we assembled a dynamical neural model of the functional core of the central complex, two structures called the protocerebral bridge and the ellipsoid body, in a simulation environment. Using non-spiking neuronal dynamics, our simulation was able to recreate in vivo behavior such as correlating body rotation direction and speed to activity bump dynamics within the ellipsoid body of the central complex. This model serves as the first step towards using idiothetic cues to track body position and orientation determination, which is critical for homing after exploring new environments and other navigational tasks.

S. C. Pickard—This work was supported by a GAANN Fellowship and National Science Foundation (Grant Number 1704366).