Autonomous vehicles rely on sensors such as LiDAR and radar to perceive their surroundings; however, these sensing systems are often developed independently of navigation applications such as Google Maps. This talk first demonstrates how location awareness, enabled by static street-topology maps, can be leveraged to adapt LiDAR and radar sensing. A live demonstration of a location-aware LiDAR system is included.
The second part of the talk examines the impact of position and orientation uncertainties on location-aware sensing solutions. To address this challenge, tools from stochastic optimization are employed, showing that location-aware sensing remains viable under practical levels of uncertainty. Finally, the talk outlines steps toward the commercialization of location-aware sensing and highlights promising directions for future research.
