Abstract

“This work is a synopsis of the results obtained on a project funded by the Royal Academy of Engineering (RAE) and is submitted as an end of project Technical Report to the Academy. It looks into interfacing of two models developed independently by the two collaborating partners: The University of Lincoln, UK and the Istanbul Technical University (Turkey). The University of Lincoln, in collaboration with the University of Manchester, developed the thalamo-cortico-thalamic circuit on the SpiNNaker machine during 2013-14; The Istanbul Technical University has been working on the Basal Ganglia Circuit and has generated a respectable body of work in this area over the years. The project funded by the RAE aimed to look into interfacing these two separate models of the thalamo-cortico-thalamic circuit and the Basal Ganglia circuit and implementing this on the SpiNNaker machine. The larger goal has been to initiate research into developing bespoke smart autonomous systems for search and rescue operation at the aftermath of an earthquake.
Here, we present the initial part of the work where a model of the thalamus is interfaced with the Basal ganglia circuit and implemented using PyNN on the SpiNNaker machine. Our results conform to biology where levels of Dopamine control the action or inaction of the cortical circuit via the thalamus. Thus, the technology shows promise for implementation in a decision making circuit, a desired and key attribute for autonomy. The latter part of the work looked into developing multiple channels of the circuitry and is being continued at the Istanbul Technical University as a post-project work and being implemented on the Brian platform.”