> The crash happened on a clear day with several hundred feet of visibility ahead, which means that the only way for this accident to have occurred is if Mr. Huang was not paying attention to the road, despite the car providing multiple warnings to do so.
This is very misleading, he might have seen the barrier well in time but the car could have steered towards it in the last second, like it does in several of the videos demonstrating this issue.
In the videos from others, the autopilot begins to move with the lane, but then crosses over into the barrier as it approaches. There's no way to anticipate that this is going to happen by simply looking ahead of you. There are no warning signs that something strange is about to occur. By the time you've processed that something is happening, it's well underway.
If you were to simulate this, I wonder how many people would be able to properly react? I think this would be an interesting experiment; sit people in driving simulators and have them drive around for 30 minutes or something and randomly introduce this. Even in the context of an experiment where people are likely to be more attentive due to the novelty of the experience, I'd expect to see quite slow reactions with fairly high frequency.
This is very misleading, he might have seen the barrier well in time but the car could have steered towards it in the last second, like it does in several of the videos demonstrating this issue.