New video of Tesla crash demonstrates the problem of semi-automated driving • Hacker News Board

Re: New video of Tesla crash demonstrates the problem of semi-automated driving

#711

Post by komali2 » Mon, Jan 16, 2023, 7:45 AM UTC

eric_the_read wrote: post #70

komali2 wrote: post #34

Earlier quoted context omitted.

I wonder why this is preferable to just taking the train for longer trips? Wouldn't it be nicer to sit back and let someone else do all the driving for however many hundred of you are all going to the same place anyway? Once you get past the long part of the trip you could always rent a car if it's rural or whatever.

The only train I could find that goes anywhere near my mother's house would take approximately 12 hours, with a 5 hour layover in a small town with ~nothing to do. I can drive it in maybe 1.5 hours. That's why.

Is this considered an acceptable situation to you and your community? What does your government representative say when you demand better public transit solutions?

Re: New video of Tesla crash demonstrates the problem of semi-automated driving

#712

Post by zestyping » Tue, Jan 17, 2023, 12:14 AM UTC

Tesla is primarily at fault for deceptively naming the function "Full Self Driving". It is indefensible mendacity.

I do not understand why the company has not already been sued into oblivion for an obvious lie that has killed people.

Re: New video of Tesla crash demonstrates the problem of semi-automated driving

#713

Post by Sophistifunk » Tue, Jan 17, 2023, 1:46 AM UTC

threeseed wrote: post #60

Sophistifunk wrote: post #11

This is a bug in the meat, and is not going away. The problems are well understood, thanks to decades of examples in semi-automated trains, boats, airplanes, and industrial monitoring systems. Until "self driving" is good enough that you're legally allowed to ride it drunk, it should not be allowed to do anything more complicated than radar cruise and lane-keeping.

This is a bug in Tesla's specific meat. Other self driving implementations e.g. Cruise, Waymo have LiDAR which is an additional and in my opinion necessary sensor for detecting objects. Because in this case it sure looks like a phantom breaking issue.

Tesla doesn't sell meat, the meat is what's weighing down the front seat.

Re: New video of Tesla crash demonstrates the problem of semi-automated driving

#714

Post by doctorwho42 » Wed, Jan 18, 2023, 2:16 PM UTC

shepherdjerred wrote: post #355

doctorwho42 wrote: post #267

Earlier quoted context omitted.

Perhaps we could dedicate the right most lane on every highway to becoming a train line instead of an autonomous vehicle lane? A train can move more tonnage (see people) per unit of energy (kWhr/gallon of diesel/etc.) Than almost any other mode of transportation (gosh darn sailboats and barges being so efficient!) . So why have 100's of electric motors and separate batteries to move people, when you can skip the batt…

This is a great idea, but wouldn't paving more road be significantly cheaper and quicker than building rail infrastructure? I'm all for the US pouring money into rail, but the fact is that we're _very_ bad at big infrastructure projects right now. That's not to say that we can't get better though.

The problem we have seen over the last 100 years in highway infrastructure is that demand always increases to fill capacity. The greatest example of this is LA, which has some of the largest highways in the country but still has bumper to bumper traffic.

You need to think about it on two metrics;

(1) density of transport; how many people per unit volume

(2) power to # of passengers.

On both counts trains are vastly superior to almost all other modes of transit other than boat.