Control of autonomous robots using principles of neuromodulation in ROS environment

Decision making of a vertebrate in response to the sensory signals from the environment is regulated by the neuromodulatory systems in its brain. A vertebrate's behaviors like focusing attention, cautious risk-aversion and curiosity-seeking exploration are influenced by these neuromodulators. The paper presents an autonomous robotic control approach based on vertebrate neuromodulation and its implementation on multiple open-source hardware platforms. A simple neural network is used to model the neuromodulatory functions for generating context based behavioral responses to sensory signals. The neural network incorporates three types of neurons-attention focusing cholinergic and noradrenergic (ACh/NE), curiosity-seeking dopaminergic (DA), and risk aversive serotonergic (5-HT) neurons. The implementation of the neuronal model on multiple relatively simple autonomous robots is illustrated through the interesting behavior of the robots adapting to changes in the environment. The implementation is done in open-source, open-access robotics software framework of Robot Operating System (ROS).