Self-reconfigurable robot controller

A component of a self-reconfigurable controller for a robot, wherein the self-reconfigurable controller has input/output (I/O) interfaces to enable communication with I/O peripheral devices coupled to the robot, the component comprising:

• processing circuitry;
• and
• a non-transitory computer-readable storage medium including instructions that, when executed by the processing circuitry, cause the processing circuitry to:
• register the I/O peripheral devices and associated functionalities, wherein the functionalities of the I/O peripheral devices are represented by functional semantic identifiers (FSIDs), which are non-random identifiers;
• receive a command for the robot to perform a task;
• conduct a self-awareness check to correlate functionalities to perform the task with functionalities of the I/O peripheral devices, wherein if a result of the self-awareness check is a lack of a correlation of functionalities to perform the task with functionalities of the I/O peripheral devices, provide an indication of a missing functionality of the I/O peripheral devices;
• and
• generate, based on a net of deep learning (DL) models and a result of the correlation, a target deep learning controller (TDLC) model to control the robot to perform the task, wherein the net of DL models comprise:
• a topology transformer model operable to transform an input FSID and an output FSID of the functionalities to perform the task into a sequence representing a topology of the TDLC, wherein the sequence is divided into subsequences representing respective TDLC model layers;
• a layer index classifier model operable to classify the subsequences with indexes;
• and
• a parameter sequence generator model operable to predict parameters of the TDLC model layers based on the indexes,
• wherein the TDLC model is generated layer-by-layer in a single iteration.

A robotic system, comprising:

• input/output (I/O) peripheral devices coupled to a robot;
• and
• a self-reconfigurable controller, comprising:
• I/O interfaces to enable communication with the I/O peripheral devices;
• and
• processing circuitry that is operable to:
• register the I/O peripheral devices and associated functionalities, wherein the functionalities of the I/O peripheral devices are represented by functional semantic identifiers (FSIDs), which are non-random identifiers;
• receive a command for the robot to perform a task;
• conduct a self-awareness check to correlate functionalities to perform the task with functionalities of the I/O peripheral devices, wherein if a result of the self-awareness check is a lack of a correlation of functionalities to perform the task with functionalities of the I/O peripheral devices, provide an indication of a missing functionality of the I/O peripheral devices;
• and
• generate, based on a net of deep learning (DL) models and a result of the correlation, a target deep learning controller (TDLC) model to control the robot to perform the task, wherein the net of DL models comprise:
• a topology transformer model operable to transform an input FSID and an output FSID of the functionalities to perform the task into a sequence representing a topology of the TDLC, wherein the sequence is divided into subsequences representing respective TDLC model layers;
• a layer index classifier model operable to classify the subsequences with indexes;
• and
• a parameter sequence generator model operable to predict parameters of the TDLC model layers based on the indexes,
• wherein the TDLC model is generated layer-by-layer in a single iteration.