Nengo and low-power AI hardware for robust, embedded neurorobotics

TL;DR

This paper demonstrates how Nengo facilitates rapid development of robust, embedded neurorobotic systems by integrating neural modeling with hardware deployment, addressing key challenges in sensor interfacing, signal processing, and control.

Contribution

The paper introduces Nengo's tools and workflows that simplify building and deploying neural networks on neuromorphic hardware for embedded robotics.

Findings

Successfully developed neural networks controlling a simulated rover on multiple hardware platforms.

Enhanced robotic control with neural adaptive methods during real-world tasks.

Provided open-source code for replicating the neurorobotic systems.

Abstract

In this paper we demonstrate how the Nengo neural modeling and simulation libraries enable users to quickly develop robotic perception and action neural networks for simulation on neuromorphic hardware using familiar tools, such as Keras and Python. We identify four primary challenges in building robust, embedded neurorobotic systems: 1) developing infrastructure for interfacing with the environment and sensors; 2) processing task specific sensory signals; 3) generating robust, explainable control signals; and 4) compiling neural networks to run on target hardware. Nengo helps to address these challenges by: 1) providing the NengoInterfaces library, which defines a simple but powerful API for users to interact with simulations and hardware; 2) providing the NengoDL library, which lets users use the Keras and TensorFlow API to develop Nengo models; 3) implementing the Neural Engineering Framework, which provides white-box methods for implementing known functions and circuits; and 4) providing multiple backend libraries, such as NengoLoihi, that enable users to compile the same model to different hardware. We present two examples using Nengo to develop neural networks that run on CPUs, GPUs, and Intel's neuromorphic chip, Loihi, to demonstrate this workflow. The first example is an end-to-end spiking neural network that controls a rover simulated in Mujoco. The network integrates a deep convolutional network that processes visual input from mounted cameras to track a target, and a control system implementing steering and drive functions to guide the rover to the target. The second example augments a force-based operational space controller with neural adaptive control to improve performance during a reaching task using a real-world Kinova Jaco2 robotic arm. Code and details are provided with the intent of enabling other researchers to build their own neurorobotic systems.