Nano-optoelectronics for insect brain inspired neuromorphic computing
Abstract
Autonomous vehicles (AVs) have the potential to solve important surveillance, recognisance and searchtasks in hazardous or hard-to-reach environments. However, presently their navigation and decisionmaking computing systems underperforms on robustness, efficiency and complexity compared to simpleinsects such as bees and ants. This significantly reduce their usefulness. We aim to show how III-Vnanowire (NW) based optoelectronics can solve this challenge in a superior way by on-chipimplementation of an anatomically verified, functioning model of the navigation control centre (centralcomplex) in the insect brain. Our project will demonstrate an important novel concept for creating energyefficient and fast neural networks where physical interconnects are replaced by light and nanoscalecomponents keeps energy consumption minimal.Replicating neural systems by software in standard computing hardware is certainly possible, but is manyorders of magnitudes less energy efficient than the brain. As a result, a race is presently on-going toimplement such networks with superior novel hardware solutions beyond standard CMOS. The highenergy cost in communication between artificial neurons can be dramatically reduced by using light asan optical photon only carries little energy, while the speed of the system can be significantly increasedas light moves fast. The project will be the key seeding point for a flexible, fast, efficient and robust micronscale optoelectronic platform that can be the central control and decision making system for any kind ofAV, using flexible visual or magnetic input for navigation, and with no need for GPS.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jun 17, 2020
- Source ID
- N629092012038
Entities
People
- Anders Mikkelsen
Organizations
- Lund University
- Office of Naval Research
- United States Navy