As automatic driving technology in trams becomes more widespread, safety remains the top concern for everyone. Some believe the algorithms are mature enough, while others worry about reliability in complex traffic conditions. So, in what situations can we fully trust the system to take over? In your opinion, in which scenarios does automated driving still require human supervision?
Can we fully trust the safety performance of automated trams?
👁️ 8 views💬 1 replies❤️ 0 likes
1 Replies
When discussing autonomous driving safety, there's a key point that's often overlooked: even with the most advanced algorithms, sensor performance can become the bottleneck. For example, in harsh weather conditions (heavy rain, thick fog, or strong sunlight), millimeter-wave radar may misjudge distances, and camera-based image recognition rates can drop significantly. In such cases, no matter how powerful the AI model is, it cannot accurately determine whether there are obstacles or pedestrians ahead. Therefore, the system must automatically switch to human supervision mode—or even completely suspend autonomous driving—when visibility is extremely low. This isn't an issue of algorithm maturity but rather a physical limitation. So, the question remains: When sensor data quality fails to meet safety decision-making standards, can we fully trust "autonomous" driving?