Topic Actions

Topic Search

Who is online

Users browsing this forum: Dauntless, Google [Bot] and 35 guests

What happens to all that debris?

Join us in talking discussing all things Honor, including (but not limited to) tactics, favorite characters, and book discussions.
Re: What happens to all that debris?
Post by cthia   » Fri Nov 06, 2020 11:13 am

cthia
Fleet Admiral

Posts: 14951
Joined: Thu Jan 23, 2014 1:10 pm

ThinksMarkedly wrote:
tlb wrote:PS. You and Cthia are justifying the usage of an expensive item in ways that are opposed. You are saying that an expensive item needs to be periodically taken out of use to reduce wear. He is saying that such downtime is wasting the money spent. You are concerned about MTBF and he is concerned about under-utilization. It might be that mean time between failure is not of a concern to the Captain, because the maintenance at the base handles that.


Well, if you followed the underutilisation argument, you'd say that a ship should be firing missiles and grasers, because they're expensive to build and should be utilised!

Those aren't use-it-or-lose-it arguments. Warships are supposed to be over-engineered. Cost matters, but not at the expense of survival in battle.

Meanwhile, MTBF is statistical. Ships don't go full refit every time they dock. That means they're always sailing with hardware all along the maintenance windows. Moreover, the ship probably spends far more time away from port than in port, meaning the engineering crew has more time to do maintenance while underway than while docked. It makes sense to do maintenance during those times, to avoid downtime.

It might also not be the captain's decision. This may be a fleet-wide mandate because the beancounters do get a say. This may be engineering guidance too, as some parts may not survive a year-long deployment if they're in use 100% of the time.

Think of the older compensators (and compensators for everyone else): they were only run at 80% capacity most of the time, since above that they had increased risk of failure.

I think we are short changing capacitors. Capacitors are built to handle many charge/discharge cycles. They are very robust. They can handle it. The two main enemies of capacitor failure are heat and age.

Neither of which will be a problem for missiles. Missiles will be destroyed long before the lifetime of the capacitors are reached. I suspect capacitors of the older missiles are still good. Capacitors generally outlast the devices they are used in. Barring outside interference like enemy missiles. I see no reason HV capacitors will fall short of the current specs and profile of capacitors.

So, charge all capacitors and keep them charged. Once charged, they can be kept charged with nominal power for use in missiles. You never know when a surprise attack may call for you to flush all missiles. You don't have time to charge capacitors then, and you don't need to be sending any "duds" downrange. Unless the author says plasma capacitors are quick-charge, very quick-charge, devices.

BTW, air-conditioning units are reliant on capacitors for start-up much like a ship's wedge. The AC unit's motor can not be started by power from the outlet. It needs the burst power from the capacitors.

Son, your mother says I have to hang you. Personally I don't think this is a capital offense. But if I don't hang you, she's gonna hang me and frankly, I'm not the one in trouble. —cthia's father. Incident in ? Axiom of Common Sense
Top
Re: What happens to all that debris?
Post by ThinksMarkedly   » Fri Nov 06, 2020 2:44 pm

ThinksMarkedly
Fleet Admiral

Posts: 4103
Joined: Sat Aug 17, 2019 11:39 am

cthia wrote:But I digress. You misunderstood me. Reactors "should" only be programmed in black&white, ultimately. Certain responsibilities should not fall upon the head of a computer program. It just shouldn't. Try explaining to Honor in the middle of a battle that the computer decided to cut power to the grasers.


I still disagree. They should not be programmed in such clear terms. The computers should have a lot more leeway and ensure that the humans don't make stupid mistakes. I would prefer to have a graser not fire during battle than a graser explode killing the weapons crew and blowing a hole in the armour because of a power surge.

I'm not saying that computers don't play a huge role in a reactor's design. They do. But it is more of a passive role by design, necessity, and common sense. It reminds me of the discussion "Should we take warm bodies out of our silos and leave launch decisions up to silly-cone diodes?" Or, would you really feel cozy in your bed at night knowing our nuclear power plants are completely automated? Computers are pretty busy monitoring all the affairs of the ship and the reactors. They are keeping track of surges, burnouts, blown junctions and power consumption and a host of other unmentionables and sending text messages to damage control, who alerts the Captain about damaged systems and when or if and what to take offline that isn't offline. Those are the decisions of the Captain, perhaps after recommended by the engineer when possible. But never made by the reactor. Except in severe cases where "Warning! Imminent failure of fusion bottles. Core ejection countdown initiated."


Today we definitely need a human in the control loop.

But who says the human in control is any less crazy than the computer? We reduce the likelihood of this being a problem by requiring two or more humans.

I disagree with your disagreement. All in all, I think David has done a fine job of incorporating computers into storyline. Do consider the massive computational prowess of a system that can simultaneously handle the hellish number of missiles tossed around at close to the speed of light. Point defense alone simply has to be supported by staggering processing power.


I think he's still limiting it to an extrapolation from the early 2000s and I feel you are too. The massive computational prowess you're speaking of is possible today already and takes at most a few tonnes of computers. A ship massing several hundred thousand tonnes has capacity to spare for this, not to mention 2000 more years of progress in miniaturisation and techniques.

I think what you are alluding to is a lack of breakthroughs in AI in the HV. That is an area best steered clear of by the the author. It will eat into his handwavium reserves. I've posted a countless number of times about the ultimate limitation of artificial intelligence ...


That's part of it. Yes, with a true AI, you should not see a difference between a human and an AI. See the Gordian universe (multiverse), where the president of SysGov is a complete AI who's never possessed a flesh body.

In the HV, this type of AI does not exist. But that doesn't mean dumb AIs like we have today wouldn't exist. They must, because they do today. Computers can react much faster than the humans and they should be allowed to automate things, in line with the instructions given to them by the humans. There are thousands of different subsystems on a ship that need power, a balancing act that a human simply cannot make because it's just too many inputs. The best a human can do is, as I said, select from a short menu of options and then let the computer do the rest. Think also of the discussion we keep having of a Leonard Detweiler-class ship coming into knife fighting range and firing a missile that can hit in 5 seconds: the computer must be authorised to counter that to preserve the ship.

So, no, I disagree and I think David is shortchanging the computers. But I also think it's a conscious decision of his for the HV (unlike the Gordian Division) and since it is his playground, he can set the rules. That's why I don't think there's much further we can discuss here: we have to accept the ground rules and move on.
Top
Re: What happens to all that debris?
Post by ThinksMarkedly   » Fri Nov 06, 2020 2:51 pm

ThinksMarkedly
Fleet Admiral

Posts: 4103
Joined: Sat Aug 17, 2019 11:39 am

cthia wrote:I think we are short changing capacitors. Capacitors are built to handle many charge/discharge cycles. They are very robust. They can handle it. The two main enemies of capacitor failure are heat and age.


I agree on heat. And since those are plasma capacitors which are holding a plasma in the range of several thousand kelvin, heat applies. Therefore, one of the enemies is realised. The longer you keep them charged, the longer they remain at high temperature and thus the more prone to early failure.

I wasn't even talking about charge/discharge cycles. In fact, in my plan, they would be charged and discharged far more often. What they wouldn't do is remain charged for a prolonged period.

Neither of which will be a problem for missiles. Missiles will be destroyed long before the lifetime of the capacitors are reached. I suspect capacitors of the older missiles are still good. Capacitors generally outlast the devices they are used in. Barring outside interference like enemy missiles. I see no reason HV capacitors will fall short of the current specs and profile of capacitors.


I meant that missiles are stored uncharged. In that condition, the capacitors should not deteriorate faster than any other component in the missile.

If they were stored in charged state, I would worry.

So, charge all capacitors and keep them charged. Once charged, they can be kept charged with nominal power for use in missiles. You never know when a surprise attack may call for you to flush all missiles. You don't have time to charge capacitors then, and you don't need to be sending any "duds" downrange. Unless the author says plasma capacitors are quick-charge, very quick-charge, devices.


Exactly: unless we get information saying they can be quick-charged somehow, I'll be forced to agree that they have to keep the capacitors charged, or at least a portion of them.

It might also be that the ship needs to carry a whole lot of supplemental capacitors in storage to replace them during long-term deployments.
Top
Re: What happens to all that debris?
Post by Jonathan_S   » Fri Nov 06, 2020 4:16 pm

Jonathan_S
Fleet Admiral

Posts: 8269
Joined: Fri Jun 24, 2011 2:01 pm
Location: Virginia, USA

ThinksMarkedly wrote:Today we definitely need a human in the control loop.

But who says the human in control is any less crazy than the computer? We reduce the likelihood of this being a problem by requiring two or more humans.

Speaking of crazy humans in the loop, go back to the Battle of Guadalcanal where IIRC it was the officer in charge of damage control officers who was so concerned about an eronious circuit breaker trip cutting off power to a combat critical system that he'd issued standing instructions than some of the breakers be "locked in" so they couldn't trip.


The commendable goal of keeping power to critical combat systems had the unfortunate side effect, when the ship came under heavy caliber fire, of magnifying, instead of isolating, the electrical damage. Instead of allowing circuit breakers to cut off shorted systems, even if they were combat critical, instead that electrical damage brought down he ship's entire electrical system, so all systems were without electrical power for the remainder of the fight! (Good thing Washington was there to ambush the Kirishima in turn and end the fight before she was able to inflict even more damage on the largely combat ineffective South Dakota)

The impact would have been lower had those "black and white" automated safety systems been allowed to activate and cut power to only localized parts of the ship. But the human in charge had them preemptively overridden; in what in retrospect was a very unfortunate decisions (causing an outcome that should have been quite foreseeable)
Top
Re: What happens to all that debris?
Post by ZVar   » Fri Nov 06, 2020 10:05 pm

ZVar
Lieutenant Commander

Posts: 114
Joined: Fri Jul 04, 2014 1:45 pm

ThinksMarkedly wrote:Today we definitely need a human in the control loop.


Do we really? Let's look at our electrical infrastructure. Almost all black/brown outs are due to human error and the "AI", I.E. the safety systems in place, trip to prevent worse damage in a larger area.

I mean yes, never make something 100% self sustaining if for no other reason we'll loose institutional knowledge of how to operate said systems, but on the same token we no longer need 100's of operators peering at gauges and dials at each electrical substation.
Top
Re: What happens to all that debris?
Post by tlb   » Fri Nov 06, 2020 10:56 pm

tlb
Fleet Admiral

Posts: 3854
Joined: Mon Sep 03, 2012 11:34 am

ThinksMarkedly wrote:Today we definitely need a human in the control loop.

ZVar wrote:Do we really? Let's look at our electrical infrastructure. Almost all black/brown outs are due to human error and the "AI", I.E. the safety systems in place, trip to prevent worse damage in a larger area.

I mean yes, never make something 100% self sustaining if for no other reason we'll loose institutional knowledge of how to operate said systems, but on the same token we no longer need 100's of operators peering at gauges and dials at each electrical substation.

The author put the following words in the mouth of one of the characters in Echoes of Honor, chapter 18:
Smith supposed it was inevitable—human beings, being human beings—that the new concept would have its critics, and some of the criticisms were no doubt valid. He did tend to get just a bit pissed off with the ones who caterwauled about what a heavy reliance the new design placed on the ship's computers, though. Of course it put a heavy demand on them . . . and anyone but an idiot knew that had always been the case. Human beings could do many of the things their electronic minions normally took care of for them, but they could do very few of those things as well—or in anything like the same amount of time—as their computers could. And there were any number of things people couldn't do without computers. Like navigate a starship. Or run a fusion plant. Or any one of a zillion other absolutely essential, extremely complex, time-critical jobs that always needed doing aboard a warship. It probably made sense to minimize total dependency on the computers and AI loops as much as possible, but it simply couldn't be entirely eliminated. And as long as he had an intact electronics shop, with one machine shop to support it, and power, and life support, Scooter Smith could damned well build any replacement computer his ship might need.
Top
Re: What happens to all that debris?
Post by ThinksMarkedly   » Sat Nov 07, 2020 12:07 am

ThinksMarkedly
Fleet Admiral

Posts: 4103
Joined: Sat Aug 17, 2019 11:39 am

ZVar wrote:Do we really? Let's look at our electrical infrastructure. Almost all black/brown outs are due to human error and the "AI", I.E. the safety systems in place, trip to prevent worse damage in a larger area.

I mean yes, never make something 100% self sustaining if for no other reason we'll loose institutional knowledge of how to operate said systems, but on the same token we no longer need 100's of operators peering at gauges and dials at each electrical substation.


Another even closer example is flying an aeroplane. All the most modern civilian airlines are flown by wire, meaning that there is no mechanical cable from the yoke to the control surfaces. And that's just on Boeing, that still has a yoke in the first place (Airbus has a joystick).

Pilots are almost not needed. On the vast majority of flights, they could sit there in the cockpit and watch the autopilot fly the plane and land it (not sure about taking off). And on Airbus planes, the computer prevents the pilot from doing something stupid, like stalling the aeroplane. The computer was also part of the reason Sully Sullenberger managed to land on the Hudson River a few years ago.

But we still have pilots, because unexpected things happen, including automation failure. Maybe in ithe Hudson case, the computer might have been able to assess the bird strike and loss of power on both engines, communicate with an automated ATC and determine a landing location within half a second of it all happening. But we still want a human (or at least a living, sentient being that values life) to make the decision among bad options.

And anyway, the biggest problem with current types of AI is that they can only make decisions based on past training on circumstances that have happened. They are woefully inadequate with new conditions. Trying to get them to extrapolate from past training is almost "Garbage in, garbage out." Just see how many issues there have been with face detection and identification for people of racial minorities (or possibly under-represented in tech).
Top
Re: What happens to all that debris?
Post by cthia   » Sat Nov 07, 2020 10:00 am

cthia
Fleet Admiral

Posts: 14951
Joined: Thu Jan 23, 2014 1:10 pm

cthia wrote:But I digress. You misunderstood me. Reactors "should" only be programmed in black&white, ultimately. Certain responsibilities should not fall upon the head of a computer program. It just shouldn't. Try explaining to Honor in the middle of a battle that the computer decided to cut power to the grasers.
ThinksMarkedly wrote:I still disagree. They should not be programmed in such clear terms. The computers should have a lot more leeway and ensure that the humans don't make stupid mistakes. I would prefer to have a graser not fire during battle than a graser explode killing the weapons crew and blowing a hole in the armour because of a power surge.

There IS lots of leeway on a ship that size within the realm of black&white. I'm pretty certain the ship's Captain, crew, and I agree with you. I don't think we are on the same page as to what constitutes a black&white decision. The computer can handle shunting power and controlling deadly power surges. Our equipment does that today. And aboard ship the respective crew are far too busy to handle mundane black&white tasks which the computer can very well handle for itself. But you must remain mindful of designing a system that is too complicated during battle, like a system whose decisions you can't predict. You must have a system in which you know how it will react. Clear terms is exactly what you need the computer handling. You don't want engineers and technicians fighting the enemy, the ship, and the computers too. Again, what you are suggesting is "General Quarters Mode." When there is a skeleton crew and horny Jimmy Johns is up to his old tricks breaking regs in a coat closet with Becky Brown instead of manning his station. At the sound of the klaxon "Battle Mode" is selected. The more complicated a system you make the more trouble it is going to be. You are underestimating alarm bells and flashing lights. Humans love auditory and visual information. These would be engineers and technicians who are well versed in that area of the ship.

Heck, I almost started a thread long ago about the all around training an officer must have to Captain a ship. How close to an engineer must a Captain be?

These are well trained people. Let them do their jobs. I agree with the speed computers can react. Preventing power surges and cascades should be under computer control with audible warnings. But shutting down grasers because they are running too hot should not be a computer's decision. Again, we are disagreeing on what is actually black&white. The following dialogue should never occur ...

"The grasers are offline Captain because they are running hot and the computer is trying to save them."

"I'M TRYING TO SAVE YOU ME AND THE SHIP GET THEM BACK ONLINE NOW!"

cthia wrote:I'm not saying that computers don't play a huge role in a reactor's design. They do. But it is more of a passive role by design, necessity, and common sense. It reminds me of the discussion "Should we take warm bodies out of our silos and leave launch decisions up to silly-cone diodes?" Or, would you really feel cozy in your bed at night knowing our nuclear power plants are completely automated? Computers are pretty busy monitoring all the affairs of the ship and the reactors. They are keeping track of surges, burnouts, blown junctions and power consumption and a host of other unmentionables and sending text messages to damage control, who alerts the Captain about damaged systems and when or if and what to take offline that isn't offline. Those are the decisions of the Captain, perhaps after recommended by the engineer when possible. But never made by the reactor. Except in severe cases where "Warning! Imminent failure of fusion bottles. Core ejection countdown initiated."
ThinksMarkedly wrote:Today we definitely need a human in the control loop.

But who says the human in control is any less crazy than the computer? We reduce the likelihood of this being a problem by requiring two or more humans.

Their transcripts and middy crew are a start. And the experience they will accrue along the way. But you are right, like Pavel Young and Elvis Santino, the Academy may have churned out a dud. But that is what simulations are for.

Long ago I was invited to several nuclear power plants to observe a simulated trip. One of the most often asked questions is, "When there is a trip, why can't the computer perform the necessary tasks to right the wrongs?"

"Some things a computer simply cannot do."

It goes back to the insurmountable limitations of artificial intelligence that I posted on the previous page, and the fact that a computer doesn't have arms and legs. A computer cannot check the validity of its assertions. It can only warn of certain conditions. Its programming is subject to all of its systems working properly. What happens when its brain is connected to appendages of itself which are now malfunctioning and trying to make decisions about systems, which are now malfunctioning. When a circuit board goes out on your car, the computer starts making wrong decisions. This is a major problem in the admissions systems in an automobile. So, TM, you don't want a graser to fire killing crew, but what is worse is a computer making critical decisions based on inaccurate information destroying the entire ship. Under battle conditions the computers are suffering damage too. You need people in the loop to oversee the health of the decisions being made by the computer. A computer which may or may not be in it's right main-frame of mind.

cthia wrote:I disagree with your disagreement. All in all, I think David has done a fine job of incorporating computers into storyline. Do consider the massive computational prowess of a system that can simultaneously handle the hellish number of missiles tossed around at close to the speed of light. Point defense alone simply has to be supported by staggering processing power.
ThinksMarkedly wrote:I think he's still limiting it to an extrapolation from the early 2000s and I feel you are too. The massive computational prowess you're speaking of is possible today already and takes at most a few tonnes of computers. A ship massing several hundred thousand tonnes has capacity to spare for this, not to mention 2000 more years of progress in miniaturisation and techniques.

I definitely must disagree. I have a supercomputer of my own. It is a capable beast. But to achieve simply what point defense requires is impossible for today's computers. They are simply not fast enough. You would need too many mainframes operating in parallel to even attempt it. And the distance between the computers would severely limit it's capability. It is one of our design problems today. You need everything on the chip to be very close to each other. Even the short distances of components on a microchip are still too far away from each other to achieve instantaneous processing speeds. And the heat generated from our computers today would kill any stealth equation. At any rate, I'm willing to bet the HV is the only place where software lags behind computing power. See my notion about the Malign's brute force beasts.

I think what you are alluding to is a lack of breakthroughs in AI in the HV. That is an area best steered clear of by the the author. It will eat into his handwavium reserves. I've posted a countless number of times about the ultimate limitation of artificial intelligence ...
ThinksMarkedly wrote:That's part of it. Yes, with a true AI, you should not see a difference between a human and an AI. See the Gordian universe (multiverse), where the president of SysGov is a complete AI who's never possessed a flesh body.

Indeed, if it is true AI there IS NO difference except the outer shell.

The arrogance of man leads him to think he can create life, but he doesn't believe that an entity could have already beaten him to it. shrug

ThinksMarkedly wrote:In the HV, this type of AI does not exist. But that doesn't mean dumb AIs like we have today wouldn't exist. They must, because they do today.

They exist in the HV as well. Remember the AI the Peeps used to determine Cutworm's likely targets? Impressive. Honor the Alpha was beaten by an AI. LOL

ThinksMarkedly wrote:Computers can react much faster than the humans and they should be allowed to automate things, in line with the instructions given to them by the humans. There are thousands of different subsystems on a ship that need power, a balancing act that a human simply cannot make because it's just too many inputs. The best a human can do is, as I said, select from a short menu of options and then let the computer do the rest. Think also of the discussion we keep having of a Leonard Detweiler-class ship coming into knife fighting range and firing a missile that can hit in 5 seconds: the computer must be authorised to counter that to preserve the ship.

Computers are perfect for running black&white errands. It is your perfect errand boy. Allocate to silly Caesar what is Caesar's. But leave the gray areas up to gray matter.

ThinksMarkedly wrote:So, no, I disagree and I think David is shortchanging the computers. But I also think it's a conscious decision of his for the HV (unlike the Gordian Division) and since it is his playground, he can set the rules. That's why I don't think there's much further we can discuss here: we have to accept the ground rules and move on.

We've reached an impasse.

Son, your mother says I have to hang you. Personally I don't think this is a capital offense. But if I don't hang you, she's gonna hang me and frankly, I'm not the one in trouble. —cthia's father. Incident in ? Axiom of Common Sense
Top
Re: What happens to all that debris?
Post by cthia   » Sat Nov 07, 2020 11:02 am

cthia
Fleet Admiral

Posts: 14951
Joined: Thu Jan 23, 2014 1:10 pm

ThinksMarkedly wrote:
cthia wrote:I think we are short changing capacitors. Capacitors are built to handle many charge/discharge cycles. They are very robust. They can handle it. The two main enemies of capacitor failure are heat and age.


I agree on heat. And since those are plasma capacitors which are holding a plasma in the range of several thousand kelvin, heat applies. Therefore, one of the enemies is realised. The longer you keep them charged, the longer they remain at high temperature and thus the more prone to early failure.

I wasn't even talking about charge/discharge cycles. In fact, in my plan, they would be charged and discharged far more often. What they wouldn't do is remain charged for a prolonged period.

Neither of which will be a problem for missiles. Missiles will be destroyed long before the lifetime of the capacitors are reached. I suspect capacitors of the older missiles are still good. Capacitors generally outlast the devices they are used in. Barring outside interference like enemy missiles. I see no reason HV capacitors will fall short of the current specs and profile of capacitors.


I meant that missiles are stored uncharged. In that condition, the capacitors should not deteriorate faster than any other component in the missile.

If they were stored in charged state, I would worry.

So, charge all capacitors and keep them charged. Once charged, they can be kept charged with nominal power for use in missiles. You never know when a surprise attack may call for you to flush all missiles. You don't have time to charge capacitors then, and you don't need to be sending any "duds" downrange. Unless the author says plasma capacitors are quick-charge, very quick-charge, devices.


Exactly: unless we get information saying they can be quick-charged somehow, I'll be forced to agree that they have to keep the capacitors charged, or at least a portion of them.

It might also be that the ship needs to carry a whole lot of supplemental capacitors in storage to replace them during long-term deployments.

I was allowing for the fact that it appears heat is not a problem for plasma capacitors. The reverse might be true.

And true, capacitors are always a dangerous technology. Oftentimes capacitor failures are made known by explosions. But I think it is the nature of the beast. David can easily wave it all away with instant charge capabilities.

Son, your mother says I have to hang you. Personally I don't think this is a capital offense. But if I don't hang you, she's gonna hang me and frankly, I'm not the one in trouble. —cthia's father. Incident in ? Axiom of Common Sense
Top
Re: What happens to all that debris?
Post by Jonathan_S   » Sat Nov 07, 2020 12:07 pm

Jonathan_S
Fleet Admiral

Posts: 8269
Joined: Fri Jun 24, 2011 2:01 pm
Location: Virginia, USA

cthia wrote:These are well trained people. Let them do their jobs. I agree with the speed computers can react. Preventing power surges and cascades should be under computer control with audible warnings. But shutting down grasers because they are running too hot should not be a computer's decision. Again, we are disagreeing on what is actually black&white. The following dialogue should never occur ...

"The grasers are offline Captain because they are running hot and the computer is trying to save them."

"I'M TRYING TO SAVE YOU ME AND THE SHIP GET THEM BACK ONLINE NOW!"

Which is why I speculate that those kind of situations the computers would alert the human operators and grant them the decision; but also have fallback decision making authority if the situation worsens and nobody has provided human input.

In this graser scenario, if nobody in the on-mount crews, the damage control, tactical, or general engineering has been able to say whether it's critical to keep the failing graser in operation then things are so messed up that cutting power to it is the least of the ship's problems (but letting it fail in some catastrophic manner might still make things much worse)
Top

Return to Honorverse