Intelligent modular robotic apparatus and methods

A method for use in expanding the functionality of a robotic apparatus, the method comprising:

• establishing a connection to a modular robotic component via a data interface;
• receiving a network image file comprising a first mapping of a first artificial neural network for the modular robotic component;
• and
• applying a second mapping of a plurality of neurons from a modeled artificial neural network to a plurality of inputs of the first mapping, the second mapping being applied to enable control of the modular robotic component.

A non-transitory computer readable apparatus configured to store at least one computer program thereon, the at least one computer program comprising a plurality of instructions configured to, when executed:

• load one of a plurality of simulated neural network image files onto a neuromorphic apparatus each simulated neural network image file specifying an artificial neural network corresponding to a finite state machine in a particular configuration;
• establish connections to a plurality of robotic components via at least one data interface;
• detect an activity state of the simulated neural network related to at least one of the plurality of robotic components;
• and
• transmit a command to the at least one of the plurality of robotic components, the command being based at least in part on the activity state of the simulated neural network.