Waymo 的 AI 与基础模型负责人 Vincent Vanhoucke，谈到当车辆上的激光雷达、雷达和摄像头彼此给出不一致信息时，谁拥有最终裁决权： “并不是它们在投票。这是一种对不同信息的融合。我喜欢用这样一个例子来解释：你的左眼基本上会告诉大脑，你的鼻子在右边；你的右眼会告诉大脑，你的鼻子在左边。并不是某一只眼睛会赢得这场‘争论’。你大脑所做的，是把这些略有冲突的信息融合起来，从而获得对眼前场景的整体理解。 安全在很大程度上来自冗余。真正的安全，源于从不同信息来源获取数据，对任何单一来源都不百分之百信任，而是根据各类信息线索去合并证据。你加入的、来自不同来源的信息越多，对现实状态的整体判断就越准确。不同传感器之间发生的这种融合，正是实现高度安全感知系统的关键所在。” Waymo’s AI and foundation model lead, Vincent Vanhoucke, on who has the final authority when LiDAR, radar, and cameras on the vehicle disagree: “It’s not like they’re voting. It’s a merger of the different information. The example I like is this: your left eye basically tells your brain that your nose is on the right; your right eye tells your brain your nose is on the left. It’s not like one eye is going to win that contest. You basically fuse in your brain the different information that’s slightly conflicting, but that actually gives you a global picture of the scene in front of you. Safety is a lot about redundancy. Safety really comes from taking different sources of information, never entirely trusting them 100%, and merging the evidence based on the different hints of information you get. The more information you add that comes from different sources, the more you improve the overall estimate of your belief. The fusion that happens between the different sensors is really the crux of enabling a very safe perception stack.”