LiDAR and Optical Filters – Helping Autonomous Vehicles See More Clearly
“What’s a steering wheel?” At the present time this would be a very strange question to hear asked from anyone who has driven, ridden in, or even seen a car but in a couple of decades this may not seem so unusual. The evolution of increasingly affordable and capable sensing and imaging systems combined with the desire to create safer, more efficient transportation systems is driving the development of autonomous vehicles (pun intended). LiDAR is a key technology that will eventually help carry this growth through to “Level 5” autonomy : no steering wheels, no brake pedals, no human intervention in driving.
Many high-end (and increasingly middle and lower end) cars already have some advanced driver-assistance systems (ADAS) to aid with collision detection/avoidance and lane departure warnings, providing “Level 1 to Level 2” autonomy, with some proto-type/pre-production vehicles, such as the Audi 8, capable of fully hands-free operation (“Level 3”). Some companies like Google/Waymo and Volvo are skipping the interim steps and working on development of fully autonomous “Level 5” vehicles in the belief that consumer traction will only be gained with fully automated vehicles, not a hybrid approach. It is too late to close the door as the “autonomous horse” is already out of the barn.
Automotive autonomy is fundamentally about vehicles responding automatically – navigating, braking, accelerating, steering – to cues from their environment; whether other vehicles, road conditions, pedestrians or other potential, and unpredictable obstacles or hazards. The old adage “you can’t manage what you don’t measure” applies to automotive autonomy; vehicles need to measure their environments and their place with in these dynamically changing scenes by accurately sensing, imaging, and detecting their conditions and surroundings in order to manage their way safely through the entire road infrastructure and surrounding environs.
There are many sensor technologies that can contribute to creating a dynamic 3D map of a vehicle and its surroundings but the three key technologies that provide guidance for automation of steering, accelerating and braking (that is, on-road driving) are RADAR, cameras, and, more recently LiDAR.