The Need for Raw Data Sensor Fusion in ADAS

This White Paper discusses the necessity of a raw data sensor fusion-based perception solution for advanced driver assistance systems (ADAS) and autonomous driving (AD) systems. It outlines the principles, configurations, and workings of sensor fusion, emphasizing its role in enhancing object detection and classification. The document explains how current sensor fusion solutions often rely on individual sensor capabilities, which can lead to performance limitations. It details three main configurations of sensor fusion: complementary, competitive, and cooperative, each serving distinct purposes in improving data accuracy and reliability. Additionally, the paper introduces LeddarVision™, a commercially available perception solution that utilizes raw data sensor fusion to create a comprehensive 3D RGBD environment model, thereby enhancing the functionality of ADAS and AD systems. The advantages of this solution include superior performance, flexibility, and compatibility with various sensor architectures, which are critical for the safety and reliability of autonomous vehicles.