Obstacle Avoidance and Target Acquisition for Robot Navigation Using a Mixed Signal Analog/Digital Neuromorphic Processing System

Abstract

Neuromorphic hardware emulates dynamics of biological neural networks in electronic circuits offering an alternative to the von Neumann computing architecture that is low-power, inherently parallel, and event-driven. This hardware allows to implement neural-network based robotic controllers in an energy-efficient way with low latency, but requires solving the problem of device variability, characteristic for analog electronic circuits. In this work, we interfaced a mixed-signal analog-digital neuromorphic processor ROLLS to a neuromorphic dynamic vision sensor (DVS) mounted on a robotic vehicle and developed an autonomous neuromorphic agent that is able to perform neurally inspired obstacle-avoidance and target acquisition. We developed a neural network architecture that can cope with device variability and verified its robustness in different environmental situations, e.g., moving obstacles, moving target, clutter, and poor light conditions. We demonstrate how this network, combined with the properties of the DVS, allows the robot to avoid obstacles using a simple biologically-inspired dynamics. We also show how a Dynamic Neural Field for target acquisition can be implemented in spiking neuromorphic hardware. This work demonstrates an implementation of working obstacle avoidance and target acquisition using mixed signal analog/digital neuromorphic hardware.

A 0.086-mm<formula> <tex>$^2$</tex> </formula> 12.7-pJ/SOP 64k-Synapse 256-Neuron Online-Learning Digital Spiking Neuromorphic Processor in 28nm CMOS

“…Information processing is entirely eventdriven, leading to sparse low-power computation [5]. This two-fold paradigm shift could lead to new bio-inspired and power-efficient neuromorphic computing devices, whose sparse event-driven data acquisition and processing appear to be particularly suited for distributed autonomous smart sensors for the IoT relying on energy harvesting [1], [3], closed sensorimotor loops for autonomous embedded systems and robots with strict battery requirements [7], [8], brain-machine interfaces [9], [10] and neuroscience experimentation or biohybrid platforms [11], [12]. However, despite recent advances in neuroscience, detailed understanding of the computing and operating principles of the brain is still out of reach [13].…”

Adaptive motor control and learning in a spiking neural network realised on a mixed-signal neuromorphic processor

“…Some of the latter devices were optimised in their small form factor and ultra low power consumption for robotic, i.e. real-time and embedded, applications [33], [24], [22].…”

“…This led us to a first neuromorphic realisation of a PI-controller. While several perception and robotic architectures were introduced recently using mixed-signal neuromorphic devices [3], [5], [17], [22], their applicability for closed-loop motor control has not been shown yet. The analog, subthreshold circuits are known to suffer from mismatch of computing elements and limited number of parameters to configure spiking neural networks [27], making their use for motor control -which requires fast and precise feedback -challenging.…”

A Braitenberg Vehicle Based on Memristive Neuromorphic Circuits

“…The results show that the vehicles with different connections between front sensors and rear wheels exhibit entirely different behaviors under the same stimulus. This famous thought experiment has inspired the development of autonomous robots that are capable of avoiding obstacles and path tracking as well as navigation . In spite of much success in achieving autonomous robots based on this vehicle model, the processors of the smart robots are still based on von Neumann architectures .…”