EBRAINS 2.0 01/01/2024-31/12/2026
The DEMONSTRATE task will develop three demonstrators illustrating the contribution of modelling and simulation services to EBRAINS workflows.
Demonstrator 1 will illustrate a use case in neurorobotics integrating spiking neural networks (SNNs) and AI modules to construct cognitive brain scaffolds. Demonstrator 2 will illustrate modelling of multiscale cause-effects across molecular/subcellular, cellular andmicrocircuit scales. Demonstrator 3 will use WP2 data to highlight the application of virtual brain twins to
epilepsy surgery.
Demonstrator 1. We will simulate active object manipulation using a biologically realistic hybrid controller operating in a closed loop. The controller will integrate an SNN cerebellar circuit performing sensorimotor prediction based on error-driven learning and will be assisted by AI algorithms (such as deep convolutional and recurrent neural networks) for object recognition. The SNN will be implemented with BSB and NEST and connected through the thalamus to an abstract representation of the cerebral cortex and basal ganglia.
