Episode 0022: Sensor Fusion

🎙️ Episode 22: Sensor Fusion in Autonomous Vehicles 🎙️

In this informational episode of DRIVEN, hosted by Paul Perrone, we delve into the intricate process of sensor fusion in autonomous vehicles. Sensor fusion involves resolving whether objects detected by different sensors, or the same sensor over time, are the same or distinct. This episode provides a comprehensive overview of how autonomous vehicles combine data from multiple sensors to accurately perceive their environment and make informed decisions.

Key Topics:

- Overview of different sensors used in autonomous vehicles (e.g., LiDAR, radar, cameras)

- Attributes and characteristics of detected objects (e.g., width, height, speed, direction)

- Temporal fusion: integrating data from the same sensor over time

- Spatial fusion: combining data from different sensors around the vehicle

- Pools of objects fused from vehicle sensors and external sources (cloud, V2V, V2I communications)

- Methods for comparing and merging distinct objects during fusion

- Culling of stale objects to maintain relevant data

- The role of confidence in determining the relevance and assuredness of detected objects

📲 Catch this enlightening episode on your favorite podcast platform and learn the details that guide autonomous driving!