- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: Uber Halts Autonomous-Car Testing After Fatal Arizona Crash
Posted on 3/19/18 at 1:39 pm to Volvagia
Posted on 3/19/18 at 1:39 pm to Volvagia
quote:
How’d you feel about a car that was manually driven but would disregard or modify the human input depending on what the computer thought was logical in that situation?
Because we’ve been flying that way for decades.
A coordinated system of human drivers with automated safeguards is probably the way of the future for the foreseeable time frame.
Hell, it's already done that way in most modern vehicles with lane warning systems and adaptive cruise control.
0
Posted on 3/19/18 at 1:40 pm to shel311
quote:
Who did?
We're going to need a lot more details before we can determine this.
uber did.
Posted on 3/19/18 at 1:42 pm to Volvagia
quote:
And this has what to do with the fact that ultimately, planes are flying around at a third the speed of sound driven by computer code.
That’s been your expressed issue, right? Prospect of buggy code making it actually more dangerous than human driving it?
And no, they aren’t dodging birds you fool.
But imagine for a tricky landing with lots of variable crosswinds the computer stuttered on its delivery.
Then there is a big thing you could hit: the ground.
Is this serious? Is someone really using aviation autopilot as a comparison for autonomous cars? The logical fails in this are too numerous to count.
We'll just go with false-analogy fallacy at this point and pretend someone really didn't attempt such a thing.
Posted on 3/19/18 at 1:44 pm to TH03
quote:
And if the waymo car isn't at fault, the data is meaningless. If someone rear ends a driverless cars, the fact that it is driverless has nothing to do with it.
But if driverless cars get rear-ended more often than cars with a human driver, that's a problem for driverless cars to address.
If you're involved in 5 wrecks in a year, and none of them are your fault, your insurance is still likely to go up.
Posted on 3/19/18 at 1:51 pm to notsince98
quote:Uber didn't.
uber did.
Posted on 3/19/18 at 1:52 pm to slackster
quote:Solution: More driverless cars, or more to the point, less human drivers.
But if driverless cars get rear-ended more often than cars with a human driver, that's a problem for driverless cars to address.
Posted on 3/19/18 at 1:52 pm to slackster
ive been in the software development business for 30+ years and currently responsible for helping to keep a system running that serves 5 million users that requires a slew of weekly patches to keep running.
i wouldn't trust a self-driving car to successfully navigate a pothole ridden street in NOLA without breaking an axle much less turn one loose on the general population.
the number of things that can go wrong here are astounding.
i wouldn't trust a self-driving car to successfully navigate a pothole ridden street in NOLA without breaking an axle much less turn one loose on the general population.
the number of things that can go wrong here are astounding.
This post was edited on 3/19/18 at 1:53 pm
Posted on 3/19/18 at 1:54 pm to CAD703X
I'm not reading through this all.
Do we have road ragin' autonomous vehciles yet or what?
Do we have road ragin' autonomous vehciles yet or what?
Posted on 3/19/18 at 1:58 pm to notsince98
quote:
Is this serious? Is someone really using aviation autopilot as a comparison for autonomous cars? The logical fails in this are too numerous to count.
No, I am not.
Try reading it again, but this time not assuming you know everything, and maybe you’ll understand.
The bulk of my content was how commercial airline pilots have essentionally zero direct controls of their planes.
Ever.
It’s all interpreted by computer, which reserves the ability to disregard or modify the commands.
Which IS highly relevant with CAD’s claimed problems to autonomous driving cars.
Obviously it’s made to work however.
This post was edited on 3/19/18 at 2:00 pm
Posted on 3/19/18 at 2:00 pm to CAD703X
I can’t imagine self driving cars running around New Orleans
Posted on 3/19/18 at 2:00 pm to Volvagia
quote:
Which IS highly relevant with CAD’s claimed problems to autonomous driving cars.
no, its not. the 'unknown' factors are essentially wind speed, rain, other airplanes and maybe an active volcanoic cloud or 2.
have you ever driven a car on actual real-life roads?
there are no universal roads; every road is a rubik's cube of challenges for a car to figure out what the hell to do. y'all gonna bitch that humans are imperfect well that is NOTHING compared to the imperfect roads these cars will be driving on.
Posted on 3/19/18 at 2:02 pm to CAD703X
quote:Because they were made by humans. Just wait until the robots make the roads!!!!
y'all gonna bitch that humans are imperfect well that is NOTHING compared to the imperfect roads these cars will be driving on.
Posted on 3/19/18 at 2:04 pm to shel311
quote:
Because they were made by humans. Just wait until the robots make the roads!!!!
shel311, pray tell what does your robot-car do if a critical sensor fails or otherwise incapacitated (aka mud splashing over the sensor, etc.) pick one:
1. stop immediately (that will be awesome for the traffic behind it)
2. attempt to safely pull over (even though the sensor is out)?
3. continue in a straight line until....?
4. FILL IN YOUR OWN
This post was edited on 3/19/18 at 2:08 pm
Posted on 3/19/18 at 2:12 pm to CAD703X
quote:I don't know the answer as I never claimed that I had all or any answers, but I imagine the companies pouring billions of dollars into the technology aren't planning to fold the entire concept of driverless cars because of mud on sensors.
shel311, pray tell what does your robot-car do if a critical sensor fails or otherwise incapacitated (aka mud splashing over the sensor, etc.) pick one:
But the answer to your question right now as it pertains to the OP is I believe the autonomous function is shut off with a warning sent to the driver that the driver must take back control.
This post was edited on 3/19/18 at 2:13 pm
Posted on 3/19/18 at 2:13 pm to shel311
quote:
But the answer to your question right now as it pertains to the OP is I believe the autonomous function is shut off with a warning sent to the driver so the driver can take over control.
so you're finally admitting an "imperfect human" driver is required for these flawless robot cars?
This post was edited on 3/19/18 at 2:14 pm
Posted on 3/19/18 at 2:14 pm to CAD703X
quote:Right now? Yes, like I have all along, nothing changed in my argument. Unfortunately, no gotcha moment for you.
so you're finally admitting a human driver is required?
A human driver was also in the car in the OP that hit and killed the bike rider, so...
Posted on 3/19/18 at 2:19 pm to shel311
i posted like 5 pages ago i was 100% in agreement for computerized DRIVER ASSISTANCE but i draw the line at car software that lets the drivers doze off/zone out/etc like the tesla and uber accidents. THAT IS A SOFTWARE FLAW.
you seem to think if a driver isn't required to be engaged that they'll have some kind of moral duty to do so when time after time its shown that's not the case.
so the gotcha moment is an ALERT HUMAN is the key here and if that's required, then AUTONOMOUS cars will never happen.
eta you even used the word autonomous so who's backtracking now?
you seem to think if a driver isn't required to be engaged that they'll have some kind of moral duty to do so when time after time its shown that's not the case.
so the gotcha moment is an ALERT HUMAN is the key here and if that's required, then AUTONOMOUS cars will never happen.
eta you even used the word autonomous so who's backtracking now?
This post was edited on 3/19/18 at 2:22 pm
Posted on 3/19/18 at 2:21 pm to CAD703X
quote:Based on where the technology is right now, that is the requirement right now. That does not mean it'll be the requirement for all time though.
so the gotcha moment is an ALERT HUMAN is the key here and if that's required, then robot cars will never happen.
Posted on 3/19/18 at 2:33 pm to CAD703X
I'm going to need you to calm down.
Posted on 3/19/18 at 2:40 pm to CAD703X
quote:
there are no universal roads; every road is a rubik's cube of challenges for a car to figure out what the hell to do.
with correct instrumentation (such as LIDAR and IMUs for starters), localized disturbances on the road could be (and eventually will be basically perfectly) accounted for
ETA: The instrumentation is already pretty capable of accomplishing these tasks. The algorithms used to interpret the data they give in a timely manner is ever-evolving though. The time-sequencing of information through a CAN bus and ultimately the Kalman filtering or similar technique to interpret the data and map the correct road conditions is where most of the research is right now, at least at Auburn that was the case.
This post was edited on 3/19/18 at 2:44 pm
Page 8 of 14
Popular
Back to top
Follow TigerDroppings for LSU Football News