Self-Recovery System for LiDAR-Based Autonomous Driving

10.29428/9789860544169.201801.0007

Ruo-Yun Wang；Shih-Fan Chou；Ai-Chun Pang

V2V communication ； self-driving car ； LiDAR

NCS 2017 全國計算機會議

2017（2018 / 01 / 01）

36 - 40

英文

For the past few years, many well-known companies and automakers have put a lot of effort into developing self-driving cars (SDCs). The SDC is undoubtedly a future trend for driving and needs various kinds of sensors to detect the nearby environment. Additionally, it uses sensing data in order to plan actions. On most SDCs, LiDAR acts as the main sensor. However, we discovered that the current approach to resolving a LiDAR malfunction is very passive and causes the SDC advantage to be lost. In this paper, we propose two strategies in order to deal with LiDAR failure through V2V communication. The aim is to recover from the breakdown and safely drive the SDCs to their destination without human support. Finally, we validated the performance of both strategies through simulation and provide insights for wireless networks.