Advantages: Increasing Vehicle Range

Advanced Driving Systems are ‘Data Centers on Wheels'

Vehicle situational awareness based on camera inferences consumes massive power.

Advanced Driving Systems typically have 10 cameras operating at 60 Hz
Each image has to be analyzed by the Deep Neural Network to infer the current situation, identify hazards, and compute a safe path ahead.
MIT studies show:
One Autonomous Vehicles (AV) with Deep Neural Network inferences at 60 Hz on 10 cameras need 21.6 million inferences per hour driven. --Sudhakar et al., 2023
A simple calculation yields
- 1 million AVs driving 1 hour per day needs
- 21.6 million inferences / hour per car * 1 million cars
  = 21.6 trillion inferences / hour
Facebooks internal studies found
- Trillions of inferences per day across Facebook’s data centers --Wu et al., 2021
1 million cars use as much energy in one hour as Facebook uses globally in a single day
Data centers currently represent 1 – 1.5% of global energy consumption

RF-eye is a Key Pillar for Sustainable Advanced Driving Systems

RF-eye is a Key Pillar for Sustainable Advanced Driving Systems

RF-eye beacons negate the need to infer the identity of all the objects in every image
Only need to infer objects without a beacon
Greatly reducing computation energy consumed by the vehicle
Improves vehicle driving range by reducing the number of inferences
- while providing accurate location, and identification, of road objects
- reducing vehicle power consumption drastically
Passive RF-eye beacons require no power. Solving major challenges for Advanced Driving Systems