4 ms·
Automotive radar is clearly not up to the task of differentiating stopped cars from signs at a range far enough away to brake safely, it's true. In theory if y
by Symmetry 8y ago
Automotive radar is clearly not up to the task of differentiating stopped cars from signs at a range far enough away to brake safely, it's true. In theory if you could fit a narrow beam phased array radar on a car it could do the job but that isn't something at all in the cards. But self driving cars with lidar have enough precision to deal with this problem even if they do cost more.
Rather than showing a problem with level 4 or level 5 automony I'd argue that this, and the psychology of driving, show that level 2 and level 3 are too dangerous since they require constant driver attention without constant driver engagement. We should just abandon the idea of having those autonomy level for cars and go straight to 4 and 5. That's what Waymo and other responsible self-driving development teams are doing and it looks like it'll work.
- nradov 8y agoA better approach would be to focus on making more and more improvements to level 1 autonomy. The driver would be hands-on all the time, but the computer would intervene as necessary to prevent collisions. We already have some of that with forward collision prevention systems, blind spot warning, stability control, etc. Just keeping improving those systems by adding more sensors and authority over more vehicle controls. Level 4+ autonomy is only going to be viable on a few limited roads for decades to come. The AI and sensor challenges are tremendous. We shouldn't let that long-term goal distract us from real safety improvements that could be realized relatively quickly with minimal technical risk.