Hacker News Reader

Can you explain this more thoroughly? So the AI cannot itself resolve a situation that requires resolution within a split second, therefore it passes control (reliable, latency free?!) to a remote operations center where humans get to... assess the situation in order to remotely tell the vehicle how to react in time? I must have completely misunderstood, because I don’t see how any part of that plan is working?

View on HN

The idea is that such situations are non life threatening. The vehicle will pull over and wait for a human operator.

It's not crazy to imagine that computers could handle split second decisions better than humans but choose to bail out to the side of the road if there's something they can't handle.

There isn’t always a “side of the road” tho. (think tunnels, walls, Old Towns, busy city centres...)

Tunnels are perhaps the only "no stopping" situation where there is nowhere to pull over. Luckily, tunnels are also the absolute simplest self-driving scenario.

Tunnels have to kept clear of obstacles. If an SDC would have to stop, so would a human-driven car.

For the others, stopped cars are very common.

So long as "bail to a human" doesn't take long e.g. 5 seconds to connect to an operator, it might not need to pull aside at all.

Reply on news.ycombinator.com