HyperSpectral Technology For Autonomous Vehicles

Brian Brown, Bryan Meister, Chetan Mrutyunjaya, Jeff Kho, Paul Liu, Venkata Vaidyanathan, Vijay Narayanan

October 20, 2022

Abstract

The goal of autonomous vehicles is simple: change the world by providing vehicles that safely navigate humans to where they need to go without the human’s input. In order to provide what might be one of the clearest pictures of science-fiction coming to life since the cellular phone, automotive manufacturers and service providers need to prove that this technology will be better-than-human in terms of safety and accident avoidance. What is science-fiction today is already becoming increasing consumer demand.

This landscape examines the role HyperSpectral (HS) imaging devices will play in this effort, in what will be a collision of the imaging industry with the automotive market to enable autonomous vehicles to enter the marketplace and our lives.

HyperSpectral imaging provides a technological solution for gaps that current autonomous vehicle players have failed to achieve, and offers better-than-human safety than one most visible-light spectrum sensors can achieve. The breadth of data available when properly harnessed will be more valuable than the data currently collected from visible light sensors alone – offering a market opportunity for new players to get involved and catch-up with industry leaders in autonomous vehicles, chiefly Tesla and Waymo.

Already in use in several industries and scientific research efforts including geography, climate monitoring, agriculture, and mining, to name a few, HyperSpectral imaging sits on the outside of the autonomous vehicle market today while automotive manufacturers continue to experiment with data imaging only across a few spectrums of light. There is at present a robust ecosystem of players from auto manufacturers, vehicle services, software companies, processor/chip manufacturers, and sensor manufacturers all having a unique opportunity to capitalize on HyperSpectral imaging through research and development, vertical or horizontal integration, and automotive manufacturing.

These opportunities for investment in either hardware or software solutions to support HyperSpectral in vehicles will reward any industry player who can successfully miniaturize and implement these existing technological offerings into a vehicle cost-effectively, overcoming cost as the largest challenge that Hyperspectral currently faces as a primary inhibitor for adoption in addition to government regulations and privacy concerns.