Great post.
But regulators and lawyers is the main impediment.
But regulators and lawyers is the main impediment.
Tesla is Finished
109,276 Views |
1566 Replies |
Last: 1 yr ago by notex
Please detail your experience in the field of unmanned systems...hph6203 said:Self-driving cars are harder technology than autonomous drones. The differentiator between your drone and an autonomous car is that the technology to develop an autonomous car is much harderAg with kids said:
UAS aren't allowed to do autonomous flight out of sight of a pilot currently without a waiver from the FAA. The potential danger of a small UAS is MUCH less than a 4000 lb car driving on a busy highway surrounded by cars.
Again...tech is the easy part - I do that for a living. It's the regulators that will be the hard part.
No it's not. It may be DIFFERENT, but it's not HARDER. Explain to me your personal expertise in this field.
and when created provides much more benefit to society. They are not the same thing. Autonomous drones do not provide the same economic benefit, they do not provide the same safety benefit,and they are pretty much all risk to the regulators and marginal benefit to the producers when compared to autonomous vehicles.
The regulators don't care about that. They care about safety and risk. How many regulators for autonomy have you worked with?
Autonomous vehicles are a multi-trillion dollar opportunity for the industry, what is the addressable market for autonomous drones? BIllions? Lots more money will be put into getting an autonomous car through the regulators than for an autonomous drone. Lots.
The big aircraft manufacturers spend LOTS of money too and it still takes **** tons of years to get things certified. I mentioned the Bell 525. First flight was in 2015. Still no certified.
Part of developing the technology is making sure it is safe. That's regardless of regulators. Businesses have an interest in developing a technology that is safer than the average driver, because the average driver isn't going to say "sign me up to ride in a death trap".
Well ****ing duh.
Regulators are not what is curtailing self-driving vehicles right now, as FKA suggested, the technology is.
No...even if you keep repeating your lack of knowledge it won't make it true.
An autonomous vehicle that drives better than the average driver, which would be the goal of the business, not the regulators, is a risk to regulators because at that point the regulators are necessitating that more die than should die.
Again, you show your ignorance about unmanned systems...
This.fka ftc said:
Great post.
But regulators and lawyers is the main impediment.
Those are all pilot programs FWIW...hph6203 said:
There are self-driving cars approved by regulators on the road. No human backup. Right now. You can get in, sit in the back seat and pay for it to take you to your destination. Whether or not you actually make it there is up to the technology. They are not capable of driving in all situations and not reliable enough even in ideal conditions. Tell me again how it's regulators curtailing their development because the regulators are allowing them to operate and the company can't get them to operate reliably more than in a small area of the country in ideal conditions. They have been on the road for 4 years now.
Regulators caused this car to stall:
Regulators made an entire fleet of Waymo cars to converge on one location
https://gizmodo.com/waymos-self-driving-cars-are-mysteriously-flocking-to-a-1847862042
Regulators caused the same thing to happen to Cruise's vehicles:
https://www.cnbc.com/2022/07/01/self-driving-cars-from-gms-cruise-block-san-francisco-streets.html
It's technology stopping their proliferation. All of those cars were operating without a human driver backing them up. They failed because of technology.
YOU cannot fly an unmanned aircraft beyond visual line of sight. We can. Because we have a special waiver from the FAA to do certain operations in certain areas under certain conditions.
Their tech wasn't ready - it IS a difficult problem. But, the regulators are going to have their microscope so far up the tech companies asses it'll be ugly. And they will push back at the SMALLEST issue. As I said, regulators are risk averse.
And they don't care about the money...because the guy that signs off on it will be **** on if anything happens...so he's going to push back hard. Because HE is not getting any of that money, why should he care if THEY do?
No, it's not. It's the complexity of the task of driving which goes way beyond just steering a car. There are a lot of heuristics applied to driving that we learn and refine elsewhere, like recognizing objects, recognizing and adapting to weather conditions, understanding and applying context, and anticipating movement or behavior in others. These heuristics are not nearly as well developed in machine vision and learning as they are in people.
It comes down to an autonomous driving AI's training set being bounded, but possibilities of actual driving situations being unbounded. We can abstract all of our experience beyond driving to handle novel occurrences while driving. AI's continue to struggle with this.
It comes down to an autonomous driving AI's training set being bounded, but possibilities of actual driving situations being unbounded. We can abstract all of our experience beyond driving to handle novel occurrences while driving. AI's continue to struggle with this.
fka ftc
Joined:
Jun 25, 2020
Posts:
14,788
How long do you want to ignore this user?
ABATTBQ11 said:
No, it's not. It's the complexity of the task of driving which goes way beyond just steering a car. There are a lot of heuristics applied to driving that we learn and refine elsewhere, like recognizing objects, recognizing and adapting to weather conditions, understanding and applying context, and anticipating movement or behavior in others. These heuristics are not nearly as well developed in machine vision and learning as they are in people.
It comes down to an autonomous driving AI's training set being bounded, but possibilities of actual driving situations being unbounded. We can abstract all of our experience beyond driving to handle novel occurrences while driving. AI's continue to struggle with this.
Yeah, most people, but a ton of bad drivers. I bet if you had 100% automation vs 100% non automation, automation deaths would be near zero while the others are what they are.
Nearly zero unpredictability. In the transition, there can be roads similar to HOV where autonomous vehicles drive.
That's another good point. Non-starter until there is a level playing field.
There is no ATC in car world. Autonomous will require ID of type, size and path of all the other vehicles.
Self driving cars are never happening. Personal air travel will happen much sooner.
Terrestrial transportation of people is sunsetting.
There is no ATC in car world. Autonomous will require ID of type, size and path of all the other vehicles.
Self driving cars are never happening. Personal air travel will happen much sooner.
Terrestrial transportation of people is sunsetting.
I had free use of Waymo back when they were beta testing. I took maybe 10 rides with them. They always had a human in the driver's seat just in case. I'm not sure if they still do or not.
Premium said:ABATTBQ11 said:
No, it's not. It's the complexity of the task of driving which goes way beyond just steering a car. There are a lot of heuristics applied to driving that we learn and refine elsewhere, like recognizing objects, recognizing and adapting to weather conditions, understanding and applying context, and anticipating movement or behavior in others. These heuristics are not nearly as well developed in machine vision and learning as they are in people.
It comes down to an autonomous driving AI's training set being bounded, but possibilities of actual driving situations being unbounded. We can abstract all of our experience beyond driving to handle novel occurrences while driving. AI's continue to struggle with this.
Yeah, most people, but a ton of bad drivers. I bet if you had 100% automation vs 100% non automation, automation deaths would be near zero while the others are what they are.
Nearly zero unpredictability. In the transition, there can be roads similar to HOV where autonomous vehicles drive.
It's not just the unpredictability of drivers. It's everything else. If you had 100% automation, as it stands you'd be paralyzed in instances of heavy rain, snow, fog, etc when things become very ambiguous. You also have things like pedestrians and animals that do not always behave as expected. There's also ambiguous road conditions, like whether a road is flooded and unsafe to cross or it's just a really big puddle. Normal traffic could be being redirected for an event or something else.
The issue with autonomous driving is the inability to handle all of the exceptions to everything and the deviations from normal driving caused by the environment, not just other drivers.
ETA Think of an autopilot trying to handle a bird strike and engine failure or a loss of instrumentation. It's not the other aircraft around it that are the issue. It's all of the environmental unknown unknowns.
Show me an autonomous vehicle company that says they can deploy self-driving cars, today, and the regulators are the only thing standing in their way. I'll wait.
I didn't say that, now, did I?hph6203 said:
Show me an autonomous vehicle company that says they can deploy self-driving cars, today, and the regulators are the only thing standing in their way. I'll wait.
Your reading skills are lacking.
I said tech is hard. I don't think any company is ready to go to full certification any time soon.
But, when they are, don't think it'll be even a REMOTELY quick process to get through certification. When the tech companies think they're done with the tech it will be 5-10 before they get certified. The regulators just move SLOW.
So...please take your failed strawman and put it back in your wallet for arguing with someone else who doesn't understand how the regulators work.
You jumped into the middle of an argument that was begun based upon that premise:
Not limited whatsoever by technology. Now you can make the argument that the capacity to lane keep and accelerate/decelerate is self driving technology, but I do not agree. That is his argument, but self-driving is far more complicated than just controlling inputs. It's know what inputs to utilize in certain situations. Those are technical problems, not regulatory problems.
fka ftc said:
Regarding self-driving tech... that is not limited whatsoever by technology at this point. Well, not directly.
Its limited by lawyers and liability laws. Either the tech / sw has to get to the point where they are happy, or laws are passed where automakers can release the product with a comfortable pucker of the exit valve.
Same things that limit robots, drone deliveries, etc.
Not limited whatsoever by technology. Now you can make the argument that the capacity to lane keep and accelerate/decelerate is self driving technology, but I do not agree. That is his argument, but self-driving is far more complicated than just controlling inputs. It's know what inputs to utilize in certain situations. Those are technical problems, not regulatory problems.
Great post.
But regulators and lawyers is the main impediment.
But regulators and lawyers is the main impediment.
ABATTBQ11 said:
No, it's not. It's the complexity of the task of driving which goes way beyond just steering a car. There are a lot of heuristics applied to driving that we learn and refine elsewhere, like recognizing objects, recognizing and adapting to weather conditions, understanding and applying context, and anticipating movement or behavior in others. These heuristics are not nearly as well developed in machine vision and learning as they are in people.
It comes down to an autonomous driving AI's training set being bounded, but possibilities of actual driving situations being unbounded. We can abstract all of our experience beyond driving to handle novel occurrences while driving. AI's continue to struggle with this.
While I know the regulators are not currently the bottle neck, AI will eventually make the regulators a bottleneck. We currently cannot prove the correctness of AI (specifcally deep learning used to train ML networks and the subsequent trained network used for driving). This is one of the major draw back of AI and why you have to have both diversity using both AI and conventinal computing. The compute required for self driving with conventional compute is much higher and is one of the current bottlenecks.
They're gonna love your autonomous drone replacing automobiles concept.
And the regulators will DRIVE a lot of those technical decisions. THAT is a fact.hph6203 said:
You jumped into the middle of an argument that was begun based upon that premise:fka ftc said:
Regarding self-driving tech... that is not limited whatsoever by technology at this point. Well, not directly.
Its limited by lawyers and liability laws. Either the tech / sw has to get to the point where they are happy, or laws are passed where automakers can release the product with a comfortable pucker of the exit valve.
Same things that limit robots, drone deliveries, etc.
Not limited whatsoever by technology. Now you can make the argument that the capacity to lane keep and accelerate/decelerate is self driving technology, but I do not agree. That is his argument, but self-driving is far more complicated than just controlling inputs. It's know what inputs to utilize in certain situations. Those are technical problems, not regulatory problems.
BTW, there ARE companies that are close to the tech. tuSimple had a bunch of Semis driving autonomously. Again, as a beta test with regulator waivers. Which restricted things they could do.
BTW...you made a comment awhile back that once they got the tech done it would be a short time (I think you said less than a year) until they were approved...I can assure you, that will NOT be the case.
Serious question. Do you not think these companies are already working with regulators all along the way?
Yes? And?aggievaulter07 said:
Serious question. Do you not think these companies are already working with regulators all along the way?
Do you think the aircraft companies that spend close to a decade to certify aircraft are too?
What does that matter?
I've been through this process before and work with the FAA closely in my current job to help develop future UAS rules and regulations. They take for ****ing ever to get **** done.
Ag with kids said:Yes? And?aggievaulter07 said:
Serious question. Do you not think these companies are already working with regulators all along the way?
Do you think the aircraft companies that spend close to a decade to certify aircraft are too?
What does that matter?
I've been through this process before and work with the FAA closely in my current job to help develop future UAS rules and regulations. They take for ****ing ever to get **** done.
I believe you, and wasn't looking to argue with you. Really just checking my assumption. I don't know Jack about how it works behind the scenes, but I was assuming the regulators aren't just in the dark right up until the "thing" is ready to go.
Not a problem.aggievaulter07 said:Ag with kids said:Yes? And?aggievaulter07 said:
Serious question. Do you not think these companies are already working with regulators all along the way?
Do you think the aircraft companies that spend close to a decade to certify aircraft are too?
What does that matter?
I've been through this process before and work with the FAA closely in my current job to help develop future UAS rules and regulations. They take for ****ing ever to get **** done.
I believe you, and wasn't looking to argue with you. Really just checking my assumption. I don't know Jack about how it works behind the scenes, but I was assuming the regulators aren't just in the dark right up until the "thing" is ready to go.
Sorry if I sounded confrontational.
No...the companies almost always have a regulator in the loop for most of the development and testing process. This helps to shape the way solutions are constructed. But, that doesn't mean they're quick at anything.
Trust me, I WISH it wasn't the case...they frustrate the **** out of me all the time. But, it's reality...
Nice straw man. Nobody said they'd be the next big thing. When it was pointed out there are workarounds that people have done when gas was in short supply nobody believed the person making the claim.ABATTBQ11 said:fka ftc said:I mean, I can understand you may not be aware of a concept from 1981, btu cmon man.ABATTBQ11 said:fka ftc said:
You can run a car on coal.
Combustion been reliable source of energy since cave man. But we are way beyond that now that the battery was recently invented.
What ****ing car are you going to run on coal, and exactly where are you going to get all this coal from? You gonna drive your coal car to a coal mine? Then what?
https://www.nytimes.com/1981/06/04/business/gm-displays-car-fueled-with-coal-dust.htmlMan, you guys have to be tired of being wrong by now, no?Quote:
Although two coal-burning cars were driven briefly today, Howard H. Kehrl. G.M.'s vice chairman, said commercial coal-powered vehicles would be ''products of the next century.''
''We'll see cars using coal liquids before then,'' Mr. Kehrl added. ''The economics of powdered coal versus liquids from coal or shale will determine which will be successful. We're just trying to show that we are not about to run out of energy. We can continue to have cars essentially as we know them today for hundreds of years.'' Large Coal Reserves Cited
Exactly where are these coal powered cars from 40 years ago that are the next big thing? Where can I buy one?
And WHERE is this coal going to come from?
I'm just amazed at how emotional you EV boys get.
Trump will fix it.
Just get a diesel and run it on vegetable oil...techno-ag said:Nice straw man. Nobody said they'd be the next big thing. When it was pointed out there are workarounds that people have done when gas was in short supply nobody believed the person making the claim.ABATTBQ11 said:fka ftc said:I mean, I can understand you may not be aware of a concept from 1981, btu cmon man.ABATTBQ11 said:fka ftc said:
You can run a car on coal.
Combustion been reliable source of energy since cave man. But we are way beyond that now that the battery was recently invented.
What ****ing car are you going to run on coal, and exactly where are you going to get all this coal from? You gonna drive your coal car to a coal mine? Then what?
https://www.nytimes.com/1981/06/04/business/gm-displays-car-fueled-with-coal-dust.htmlMan, you guys have to be tired of being wrong by now, no?Quote:
Although two coal-burning cars were driven briefly today, Howard H. Kehrl. G.M.'s vice chairman, said commercial coal-powered vehicles would be ''products of the next century.''
''We'll see cars using coal liquids before then,'' Mr. Kehrl added. ''The economics of powdered coal versus liquids from coal or shale will determine which will be successful. We're just trying to show that we are not about to run out of energy. We can continue to have cars essentially as we know them today for hundreds of years.'' Large Coal Reserves Cited
Exactly where are these coal powered cars from 40 years ago that are the next big thing? Where can I buy one?
And WHERE is this coal going to come from?
I'm just amazed at how emotional you EV boys get.
The strawman of EMP's led to the added idiocy of how to steal gas after the apocalypse and coal powered cars. The topic of this thread isn't antique vehicles or what to drive if you happen to survive a world ending event.
It isn't that no one believed that cars can be powered by coal, it's that cars powered by coal is completely irrelevant to the discussion at hand. I hope that helps.
It isn't that no one believed that cars can be powered by coal, it's that cars powered by coal is completely irrelevant to the discussion at hand. I hope that helps.
Alternative fuel sources for vehicles seems pretty relevant to the viability of Tesla.GAC06 said:
The strawman of EMP's led to the added idiocy of how to steal gas after the apocalypse and coal powered cars. The topic of this thread isn't antique vehicles or what to drive if you happen to survive a world ending event.
It isn't that no one believed that cars can be powered by coal, it's that cars powered by coal is completely irrelevant to the discussion at hand. I hope that helps.
An even minor EMP attack could enlighten many as to what no electricity means to modern life, and they will be reluctant to go all in.
You fancy yourself a know it all, but fail to see the forest as you piss on trees. Next time you are piloting and the airplane is in fly itself mode, pick up a book and broaden your knowledge. Just a suggestion.
Surprised that suggestion wasn't already made by the preppers. Probably even less viable than coal.
Actually, if you read the article, it IS viable and people have done it.hph6203 said:
Surprised that suggestion wasn't already made by the preppers. Probably even less viable than coal.
Not saying it's a GOOD idea, but it WILL work...
We're talking about a world that has lost electricity as an option. How are you growing enough crops to produce enough vegetable oil to convert to biodiesel to drive your car? And in such a world what kind of maniac would burn perfectly good calories to drive down the road when all the trees have been stripped of their bark just to fill people's stomachs.
I didn't say convert it to biodiesel.hph6203 said:
We're talking about a world that has lost electricity as an option. How are you growing enough crops to produce enough vegetable oil to convert to biodiesel to drive your car? And in such a world what kind of maniac would burn perfectly good calories to drive down the road when all the trees have been stripped of their bark just to fill people's stomachs.
I said use the vegetable oil as a fuel. As in just poor it in your tank...
Load up at walmart since their self checkouts won't work...
Question, have you read the book One Second After?
I've read a lot of books on the apocalypse and watched more than enough shows and movies.
Particularly the first book in the series lays out a pretty well thought out post apocalypse US.
Coal cars appear in the 2nd book or I probably would not believe it either. But then you start really thinking through things.
There may be plenty of crops after a year or two and most of the population will have died out. The folks that new before the doomsday that a car could even be modified as such will have an advantage over those who don't.
I'm not a prepper. I depend on insulin so even in EMP armageddon I am grade A fooked. But people should know history and what things are possible.
Also, EV market penetration will settle at 25%. Airliners can fly themselves. Lawyers slow progress. And cars can run on coal.
I've read a lot of books on the apocalypse and watched more than enough shows and movies.
Particularly the first book in the series lays out a pretty well thought out post apocalypse US.
Coal cars appear in the 2nd book or I probably would not believe it either. But then you start really thinking through things.
There may be plenty of crops after a year or two and most of the population will have died out. The folks that new before the doomsday that a car could even be modified as such will have an advantage over those who don't.
I'm not a prepper. I depend on insulin so even in EMP armageddon I am grade A fooked. But people should know history and what things are possible.
Also, EV market penetration will settle at 25%. Airliners can fly themselves. Lawyers slow progress. And cars can run on coal.
They're not very good about reading articles or logical arguments.Ag with kids said:Actually, if you read the article, it IS viable and people have done it.hph6203 said:
Surprised that suggestion wasn't already made by the preppers. Probably even less viable than coal.
Not saying it's a GOOD idea, but it WILL work...
I've heard the exhaust smells like a deep fryer on those.
Trump will fix it.
I don't consider myself a know-it-all but compared to you I'm feeling alright.
This thread is annoying as **** now
Quote:
I've read a lot of books on the apocalypse and watched more than enough shows and movies.
That's shocking
Featured Stories
See All
27:36
4h ago
4.4k
23:50
16h ago
9.1k
4:34
4h ago
1.4k
Game Highlights: Vanderbilt 86, No. 12 Texas A&M 84
by Cade Bickham
9:18
17h ago
3.5k
Photo Gallery: Vanderbilt 86, No. 12 Texas A&M 84
by Will Huffman
aggiez03
No rationalization exists for No. 12 Texas A&M's 86-84 loss to Vandy
in Billy Liucci's TexAgs Premium
5
aggiez03
No rationalization exists for No. 12 Texas A&M's 86-84 loss to Vandy
in Billy Liucci's TexAgs Premium
3