Starship crew

[Ahriman238]

Way back in the days of sail, large ships required hundreds of men to physically adjust the sails, perform basic cleaning and matinence, and man the guns in combat. Only a very small number of specialists, the carpenters, doctors, cooks, marines, and officers were exempt from the back-breaking labor of keeping a ship seaworthy.

In more recent times, a lot of functions have been automated, and so the bulk of a ships crew will still be involved in basic matinence, or damage control in the event of battle. Of course, you still need doctors, cooks, marines, and specialists in this and that system. And in some ships dedicated postmen, shopkeepers, lawyers and police. But a larger portion of the crew will be mechanics and electricians.

So, what sort of jobs would starship crews have? Why does it take 1,000 people to run the Enterprise, over 30,000 to run an ISD, or 11,000 for a Luna-class cruiser?

[Purple]

Well I imagine that it would depend massively on the setting and just how much automation there is and for what it is used. Depending on those factors you could be looking at anything from millions of people man handling every bit of equipment and doing all the work from port hole cleaning to loading giant shells into canons via slave labor (40K) through large crews with some automation for cleaning and stuff (Star Wars) to small groups of highly educated astronaut equivalents controlling almost completely automated starships and finally to just having one crew member per starship controlling the whole thing him self like in a Bolo.

A good example of this variety can be seen in Star Wars in fact. You have the Trade Federation that has very small crews comparatively with most of the work being automated (droid labor) and than you have the Empire where we see a lot more work being done by hand (manual gunnery for example). So really there is a myriad of factors that play into this. Technology, cultural bias, cost (it might be cheaper to do things by hand with unskilled labor than to make robots), religion etc.

[The Romulan Republic]

I don't expect any real world Starship to have a huge crew. If you look at real spacecraft, they tend to have very small crews. Apollo: three. Shuttle: seven max.? Mars mission plans talk about four or six crew. And we're perfectly capable of building deep space probes that run with no crew at all. And the more crew you have, the more the mission will cost.

Now, I can envision a scenario where a starship might be required to carry a large number of colonists or troops (probably in suspended animation of that ever becomes possible). But actual crew? Not a lot.

If I were going for realism in a sci-fi story, I'd probably write a very small crew: a captain, a pilot, a couple engineers, a couple scientists, a gunner/tactical officer for a military ship. Anyone else would be a colonist or marines in suspended animation in the cargo hold.

But these are just my guesses.

[Simon_Jester]

The fundamental problem here is the difference between a warship and a cargo ship.

A seagoing cargo ship weighing a hundred thousand tons might have a crew of, what, a few dozen? The only labor-intensive task involved in running that ship is loading and unloading its cargo, and that's handled by dock workers. Actually steering the ship, keeping track of the radio, and so on takes very few people.

A seagoing warship weighing a hundred thousand tons has a crew of thousands. Because every bit of the ship is in use for complicated tasks, there are no vast piles of goods that can be left sitting in the hold without supervision. There are more complex mechanical devices on the ship that will need to be used. There is more need for men capable of fixing things when they break, because enemy action will break things a lot more often than wind and wave and bad luck would do alone. The requirements pile up fast.

A spacegoing warship will have some of the same problems. Bits of the ship are more likely to break in combat than they would on a normal ship. Operating the warship will involve using a lot more complicated sensors and threat analysis, and maintaining a lot bigger range of equipment in different environments.

So you can't necessarily get away with the "we only need three people so one person is awake at all times" approach she might have tried...

Now, if you have fully sentient AI, or even expert systems that can do surprisingly hard jobs as well as a human could, you can cut crew back drastically- but in a real sense, all you're doing is adjusting crew size according to the number of robot crew you're using.

[Dwelf]

You also need to remember that if you want 24 hour operations you will need multiple shifts. This will rather dramatically increase the number of crew somewhere between 2 and 4 times I'd guess depending on how hard you are willing to work them and how much redundancy is acceptable.

[Number Theoretic]

Another factor that play into crew sizes is how many different jobs can be done by one person (per shift). Even without sentinent AI, it is perfectly plausible that a lot of automation can be done in a real-world spacecraft. So, for example, the computer can do some parts of the pilot's job and the navigator's job and the remaining parts that can not (or are for some other reason not) automated can be done by one person: the pilot/navigator.

Same reasoning goes for weapons in a warship: it doesn't stretch my imagination too hard that aiming and firing and maybe some target preselection can be done by a computer, which leaves the gunner just with the task of selecting targets and queuing up firing commands, which reduces the gunner's workload so much that it becomes debatable to move this task to the captain.

So, what jobs are left and how their tasks look like indeed depends heavily on automation and how it works.

[The Romulan Republic]

As an example of this, consider the original Enterprise in The Search for Spock. The standard crew was something like 400, right? But Kirk and company were able to fly it, even badly damaged, with only a crew of five.

[GrandMasterTerwynn]

Let's think this through:

A hard sci-fi starship is going to have a tiny crew and huge amounts of automation, simply because a lighter ship can get closer to the speed of light and all the shit needed to keep humans alive is fucking heavy. Then again, you're not apt to use a hard sci-fi starship as a warship simply because there aren't a whole lot of conflicts where you'd be willing to wait several decades between major battles. Even if you were willing to wait that long, the starship is likely to sit well outside the target system and spam it with missile drones controlled by high-order AI or uploaded humans anyway, because a hard sci-fi ship is going to be bloody expensive.

So . . . let's examine something more interesting.

In current wet navies, you divide the crew into two or three separate teams and they stand a total of four to seven watches in a 24 hour day. The traditional watch system has five four-hour watches and two two-hour watches. A typical merchant ship has six four-hour watches. So, whatever basic crew requirements you have must be doubled or tripled.

But that's just fine, because a starship is still likely to have significant automation. You're not going to want squishy humans anywhere near your antimatter-powered super-hyper-warp engine, after all. The gamma rays and induced radioactive decay, should something go wrong, will kill the humans from radiation poisoning in short order anyway. (Though a human is nothing more than a jury-rigged carbon-based nanotech utility fog anyway, so with sufficiently-advanced repair technology . . . well, the controlling intelligence still leaves something to be desired in stressful situations; given all the superfluous code in the software.)

They also won't be pointing the guns, because humans aren't precise enough for the long-range combat, and they may not be fast enough for the short-range engagements. Humans may have written the software that runs the fire-control and target-selection systems, but you only need a human in the loop to tell the automatics that, yes, it's okay to open the can of whoop-ass on the enemy; or no, it's time to close the can of whoop-ass.

They won't be down in the guts of the sensor systems either. A radar powerful enough to reach out and touch someone millions of kilometers away will, pretty much, be a microwave laser. And a starship's sensory systems could generate so much information that humans would be overwhelmed. Therefore, you'd have a lot of software sifting through and prioritizing the data.

Nor will they be heavily involved in steering and driving the ship. At fractional-cee velocities, they couldn't react fast enough to oncoming obstacles, and at the distance of billions or trillions of kilometers, they're not precise enough. The most you'd need is a human to direct the ship to "go here and try to do it in this much time," and let the automatics work out the details.

The maintenance and damage control functions are also better-suited to automation. A robot can be constructed with the strength and dexterity needed to quickly patch things up in environments too hazardous for humans, and you can have micro-robots for spaces too small for humans. You could even have a network of pipes containing nanotech utility-fog that will patch up the tiny problems before they become hazards.

So . . . how many people do you actually need to run a starship? Your warship's bridge will have an Officer of the Deck (OOD) and a junior Officer of the Deck (who is training to become an OOD.) The junior OOD generally mans the conn and steers the ship. There may be a quartermaster who updates the log and tracks the ship's navigation, a signalman to manage communications traffic from other ships, and a petty officer to secure the bridge and supervise who comes and goes. So, possibly five people on the bridge for day-to-day operations. You could cut this down to two by eliminating everything but the OOD and JOOD. During combat situations, you'd have an officer in tactical command of the ship (who may, or may not, be the ship's Commanding Officer,) and possibly a weapons/fire-control officer as well.

In the ship's Combat Information Center, you could have a number of people responsible for digesting the data provided by the ship's sensory systems and directing the ship's fire-control systems. They would coordinate with the bridge or whatever command center the officer in tactical command of the ship happens to be sitting. There could also be petty officers responsible for securing the CIC as well.

In the ship's engineering spaces, you might have a person monitoring the propulsion systems, a person monitoring the power systems, a person to monitor the life-support systems, one to monitor the damage-control systems, one to monitor the sensory systems, one to monitor the weapons systems, and a person responsible for supervising all of them. With enough people, there'd be someone in charge of securing this space as well . . . though you could make one or two people in charge of monitoring and directing everything engineering related.

You'd have a doctor and a couple of nurses manning the ship's infirmary. On the ship's mess decks, you could have two or three people in charge of taking orders, three or four cooks, and a couple of security people. Finally, you could throw in a handful of people responsible for the maintenance of the livable spaces of the ship, the loading and unloading of cargo, and manning shore/boarding/inspection parties.

So, you'd have a ballpark of somewhere between fifteen to sixty people per team. So a large sci-fi warship might be expected to have as little as fifty people on the whole crew, or as many as two-hundred. With extensive enough automation, you could cut this down to all of, say, twenty individuals, including the command staff. Bear in mind that this would be for a ship at least a kilometer wide and one or two or three kilometers long (here, I've chosen a ship wide enough to house a large enough electron accelerator ring to drive an x-ray laser and long enough to keep all the dangerous antimatter at a safe distance from the ship's crew.) Now you might have many more people on the ship (say a thousand marines to board and secure a ship that you rad-killed at a few light-minutes out with your x-ray laser, or pacify a small space habitat,) but they won't be crew so much as cargo.

[Purple]

Now you might have many more people on the ship (say a thousand marines to board and secure a ship that you rad-killed at a few light-minutes out with your x-ray laser, or pacify a small space habitat,) but they won't be crew so much as cargo.

I see no reason why any passengers could not be given the duties of general hands. You know, things like cleaning, low level maintenance etc. In fact, if a given warship is permanently pared with a given military unit (or any military unit as long as you can expect there to be one on board on a regular basis) I would expect this to become standard operating procedure to have the ship "crew" only consist of mission critical experts like those you described and to have any marines on board take over things like the day to day cleaning and maintenance or even emergency repair in combat. After all, it does not take that much effort to field train the grunts to use a mop and do some basic repairs like swapping out modular assemblies with minimum tools.

And that is another thing that we might want to consider. On large starships we will probably see a trend toward modular repairs for just about anything smaller than a reactor. You won't have the Chief Engineer trying to jury rig circuit boards or fix the replicators and stuff. Instead he will just go and replace the whole unit from existing stocks. And when you get to that you don't really need highly skilled crew members to run maintenance. That is a trend we are already seeing in modern military technology.

[GrandMasterTerwynn]

Now you might have many more people on the ship (say a thousand marines to board and secure a ship that you rad-killed at a few light-minutes out with your x-ray laser, or pacify a small space habitat,) but they won't be crew so much as cargo.

I see no reason why any passengers could not be given the duties of general hands. You know, things like cleaning, low level maintenance etc. In fact, if a given warship is permanently pared with a given military unit (or any military unit as long as you can expect there to be one on board on a regular basis) I would expect this to become standard operating procedure to have the ship "crew" only consist of mission critical experts like those you described and to have any marines on board take over things like the day to day cleaning and maintenance or even emergency repair in combat. After all, it does not take that much effort to field train the grunts to use a mop and do some basic repairs like swapping out modular assemblies with minimum tools.

Sure, you could have the cargo be in charge of maintaining its livable spaces. In fact, that's pretty much how they're going to spend their time in between off-ship operations and training drills. You could, indeed, even rotate them in and out of the deck and security departments; having the marines swabbing decks and touching up paint under the direction of a boatswain's mate, for instance; or cooking meals on the mess deck. However, the starship simply isn't going to need a whole lot of people to keep it running. Humans would just get in the way when it comes to maintaining vital shipboard components, or making emergency repairs in combat. About the only damage control humans would have any business doing aboard a combat starship would be firefighting, security, and search-and-rescue operations within the ship's habitable areas . . . if the ship is so badly damaged in battle that all of the sophisticated automation is no longer working, then it's probably about to become a cloud of incandescent gas and wreckage anyway.

[Dwelf]

Expert systems for damage control would be extremely hard to write due to the less than predictable nature of combat damage. You may have damage control robots but only for the simplest tasks or in the most advanced settings would I expect them to handle combat damage well unattended. Sophisticated automation also has a tendency to be fragile compared to less complex systems. I'd expect a reasonable number of people to be available for damage control in many settings.

The problem with cutting the crew down to the bare minimum. Do you trust the fire control computer to run in autistic mode when connection to your one fire control officer in the bridge is damaged. Do you instead have local fire control officers with each of the batteries in case central fire control is out of contact.

Your sensors systems will probably handle most of the target analysis and selection. There is probably far more data than a single person can handle so do you have one person taking the computer's word for things or assign a small team to double check the input incase some new stealth technology from your oponent takes advantage of a flaw in your algorithms.

You only have as much redundancy as the least redundant system if human input is required they are one of the systems. There is no point having a tripple redundant failsafe system that requires human input and only having 1 crew member and 1 station to control it.

The overhead for additional life support on a multi km long vessel might not actually be that high compared to the cost and weight of the automated systems. Overall it will come down to the cost of automation, the cost of crew and the trust in the automation. Personally I don't think tiny crews on large warships is a particularly sound choice.

[Purple]

Expert systems for damage control would be extremely hard to write due to the less than predictable nature of combat damage. You may have damage control robots but only for the simplest tasks or in the most advanced settings would I expect them to handle combat damage well unattended. Sophisticated automation also has a tendency to be fragile compared to less complex systems. I'd expect a reasonable number of people to be available for damage control in many settings.

With that much I can agree. Certainly the problem can be somewhat deminished by introducing automatization and by having has many components as possible be plug and play. But unless the universe in question has acces to Star Wars type sentient robots I think you are much closer to reality than GMT.

The problem with cutting the crew down to the bare minimum. Do you trust the fire control computer to run in autistic mode when connection to your one fire control officer in the bridge is damaged. Do you instead have local fire control officers with each of the batteries in case central fire control is out of contact.

Your sensors systems will probably handle most of the target analysis and selection. There is probably far more data than a single person can handle so do you have one person taking the computer's word for things or assign a small team to double check the input incase some new stealth technology from your oponent takes advantage of a flaw in your algorithms.

I would actually say both. Under cruising conditions you would definitively want a team of sensor technicians constantly going over the data. But in the middle of a heated battle there would simply be no time for this and you would have to trust the sensors simply to keep reaction time up.

You only have as much redundancy as the least redundant system if human input is required they are one of the systems. There is no point having a tripple redundant failsafe system that requires human input and only having 1 crew member and 1 station to control it.

What if said crew member is in the most armored and protected part of the ship with many, many redundant connections to the actual systems. Considering the sheer size of these things it is not unimaginable that the only way to sever all the connections might be to blow up one of the two end points.

The overhead for additional life support on a multi km long vessel might not actually be that high compared to the cost and weight of the automated systems.

Or you can just go cheap and only pump battle stations full of air and have people use space suits to travel betwen them. Makes for some good drama too when the boarding party realizes there ain't no air and they did not bring enough to walk several kilometers to the nearest room that does.

[Enigma]

As an example of this, consider the original Enterprise in The Search for Spock. The standard crew was something like 400, right? But Kirk and company were able to fly it, even badly damaged, with only a crew of five.

But the ship wasn't meant to be flown by such a low number of people. Definitely not for a five year mission. I'm sure a very small group of 10 experienced sailors could get an aircraft carrier underway but I doubt they can fulfill it's role effectively. But yes, Trek ships are heavily automated.

In Trek, ships with high crew complement such as the Galaxy (~1,000) and Constitution (430) (don't know the crew complements for the Sovereign, Excelsior or Nebula) is because of their roles as exploratory vessels. They are deployed on a multi-year mission, thus require a lot of specialized crew for various departments and in the case of the Galaxy, including family.

Starships with a more defined role tend to have a smaller crew complement (Oberth 5-80, Miranda 26-35, Defiant ~50) and AFAIK a limited deployment time.

[Stofsk]

Uh where are you getting estimates for the Miranda from? A Miranda should be comparable to a Connie in terms of crew complement because they have similar internal volume. If a ship like the Defiant can have 50 dudes on it then the Miranda, which is a lot bigger, should have just a bit more.

[Simon_Jester]

You only have as much redundancy as the least redundant system if human input is required they are one of the systems. There is no point having a tripple redundant failsafe system that requires human input and only having 1 crew member and 1 station to control it.

Please improve your grammar, syntax, and spelling.

Anyway, as I was going to say, there is a point in triple redundant systems with one controller if the volume/mass penalty of adding redundancy is small and the volume/mass penalty of adding another man in a separate backup control center is large. But in a sufficiently large vessel, the marginal cost of adding more people is small, which creates more incentive to add redundant control positions.

This doesn't necessarily mean adding a lot more people. It can just mean adding 'backup bridge' or 'auxiliary control' stations from which any authorized member of the chain of command can issue orders in the event that primary bridge is out of commission. If the crew is distributed throughout the hull, and the computer network is intelligently designed to make rerouting around a break practical, you don't actually need a backup command staff necessarily; you can make do with junior officers who are at least approximately trained to do the job.

Though the larger the ship gets, the more arguments there are for having a 'spare command crew,' and for having isolated points of secondary control close to the systems in question (battery fire control, stations for monitoring the reactor that are close enough for a radiation-suited man to get to the reactor and start doing things before it explodes).

The overhead for additional life support on a multi km long vessel might not actually be that high compared to the cost and weight of the automated systems. Overall it will come down to the cost of automation, the cost of crew and the trust in the automation. Personally I don't think tiny crews on large warships is a particularly sound choice.

There will be huge chunks of the ship that are not accessible to or safe for human life, especially not human life that isn't wearing special protective gear. That tends to scale back the size of the living quarters.

With that much I can agree. Certainly the problem can be somewhat deminished by introducing automatization and by having has many components as possible be plug and play. But unless the universe in question has acces to Star Wars type sentient robots I think you are much closer to reality than GMT.

Terwynn's assumption seems to be- and the advance of computer technology supports his opinion- that we are more likely to see intelligent robots of the same capability range found in Star Wars than we are to see spacegoing battleships. Fully intelligent AI capable of matching anything humans can do is at least plausible within our lifetimes; hyperdrives aren't.

So just as people normally assumed in the 1930s that the future would contain atomic power and rockets (neither of which existed at the time), Terwynn would normally assume that the future will contain computers advanced enough to do much of the thinking and hands-on work that human beings would be needed to carry out today.

In which case the main reason for having human crews in the decision loop at all is to make sure that at least one element in the decision loop is one we understand and can vet by mechanisms other than beta testing.

What if said crew member is in the most armored and protected part of the ship with many, many redundant connections to the actual systems. Considering the sheer size of these things it is not unimaginable that the only way to sever all the connections might be to blow up one of the two end points.

Sigh, we've been through this before. The volume and mass penalty of building so much armor and redundant control networks makes it silly to reject the cost of just having more people on the ship, distributed throughout the hull.

One man is a single point of failure, and piling on more redundant methods for him to control systems hits diminishing returns. Double redundancy is good, triple redundancy great, quadruple redundancy is about as far as you can go before diminishing returns. Because eventually, the odds of all the control systems being knocked out is low, but only compared to the odds of something going wrong with that one man. Which might not even be battle damage; he might be out of action with a burst appendix or freezing up in combat.

At which point it's much more profitable to invest in a backup human being to issue orders than to invest in backup commlinks to transmit those orders to the rest of the ship.

Or you can just go cheap and only pump battle stations full of air and have people use space suits to travel betwen them. Makes for some good drama too when the boarding party realizes there ain't no air and they did not bring enough to walk several kilometers to the nearest room that does.

That's not drama. That's stupidity.

[Skylon]

I don't expect any real world Starship to have a huge crew. If you look at real spacecraft, they tend to have very small crews. Apollo: three. Shuttle: seven max.? Mars mission plans talk about four or six crew. And we're perfectly capable of building deep space probes that run with no crew at all. And the more crew you have, the more the mission will cost.

I think the deceptive part of those vehicles is how much they are supported by the ground. An Apollo or Shuttle mission was backed up by hundreds of people on the ground, who had access to real-time data, and are specialized in a specific function of the craft. They aren't there because its, as you noted, too costly to put them there, and thus far our forays into space have been such that home is a quick call away - so you don't really need a specialist in each ship function there - you can just call home, and they'll tell you what switches to flip.

This gets a lot trickier if you get into an interstellar voyage, or a scenario where the ship and crew are expected to operate completely autonomously. AI will probably pick up the void left by Mission Control in reality, but on a warship you may not have time and need a specialist, who can think on their feet, right then and there.

[Ahriman238]

It seems I have, once again, expressed myself poorly in the OP. My apologies. I'll try and do better here.

Mostly, I blame myself for bringing up Star Trek which is generally an exception to what I'm thinking of. The various Enterprises have had a primarily scientific mission, and so carried large numbers of scientists, some of whom were crosstrained (If memory serves, Sulu was a physicist on the ship before he became the helmsmen. Don't ask what happened to the last guy) as well as more obscure specialities like a historian. In one episode of TNG, they were able to turn up 4 dedicated geologists among the crew, impressive when you consider how often geology is not an issue in interstellar exploration.

But how about that Star Destroyer? 30,000 ratings, not counting the stormtroopers or the 9,000 officers (is it just me, or is the officer ratio a bit high whenever they publish stats for SW ships?) I imagine you'd want a couple thousand for damage control/matinence, some more as redundant system operators. Say you need 500 at all times to monitor the reactors, assume 4 shifts we'll say another 2,000 safety crew for the reactors. Now let's say that every TIE fighter needs 100 men to service and fuel it, they wouldn't all have to be on the same shift, so 7,200. Support staff: medical, security, services (barbers, shopkeepers etc.) call it another 2,000. Let's say 1,000 divided across 4 shifts as bridge crew.

Being as generous as I can think to be, I have not quite accounted for half of the crew numbers. Ok, so maybe the Essential Guide crew figures are drawn from a hat, there to look impressive. My question to you is, what do all these spacers DO as part of their jobs? Space is a very empty enviroment, barring the occasional micro-meteor, or a very stressful form of FTL, you should hardly need much matinence which historically is most of what sailors DO.

You can pad out the numbers to an extent, a starship might need a large diplomatic staff, or a few dozen experts in xenopsychology, or someone to subdue anyone who looks out the window at the wrong time and loses it. But if you have a large crew, and there's no need to scrape space barnacles off the hull, you get large numbers of bored people. People looking for entertainment in the most deadly enviroment known to man. This is clearly not a good thing, not for morale, crew sanity or safety.

And yes, you can run a supertanker with a dozen people, because they don't require such intensive matinence or anticipate battledamage. A couple decades ago, you'd have needed 40. This is expected, as recently as WWII when the captain ordered a certain speed, his orders were relayed via telegraph to the engine room where a dozen men were required to make it so.

As an example of this, consider the original Enterprise in The Search for Spock. The standard crew was something like 400, right? But Kirk and company were able to fly it, even badly damaged, with only a crew of five.

Didn't the automation freeze the first time they tried to do something more complicated then fly to point B, then completly overload the first time they took damage?

[GrandMasterTerwynn]

Expert systems for damage control would be extremely hard to write due to the less than predictable nature of combat damage. You may have damage control robots but only for the simplest tasks or in the most advanced settings would I expect them to handle combat damage well unattended. Sophisticated automation also has a tendency to be fragile compared to less complex systems. I'd expect a reasonable number of people to be available for damage control in many settings.

If you can't build a complex automatic system that can function without meaningful human intervention for decades, then you're never leaving the Solar System . . . simply because most slower-than-light starship designs aren't going to be able to go fast enough for time dilation to be an appreciable factor. Even if there happens to be some magic trick that enables FTL travel, and we discovered it tomorrow, we'd still not be able to fully exploit it for many decades at least . . . simply because building fleets of big starships requires a whole shitload of well-developed starship-oriented industry. One might presume that increasingly sophisticated and robust automation will also continue to be developed in that timeframe.

Consider that damage control isn't necessarily easy for humans either. Which is why there are these things called procedures written to educate crewpersons in a basic set of steps to attempt to recover from a fault condition, or to diagnose what the fault condition could be. Procedures tend to come with clearly-delineated steps . . . i.e. a program for humans.

Given the present progress in computing speeds and technology, I would fully expect that we'd have damage control robots smart and fast enough to be better at diagnosing problems and rigging bypasses than humans (or, failing that, really good telepresence directed by augmented humans.)

The problem with cutting the crew down to the bare minimum. Do you trust the fire control computer to run in autistic mode when connection to your one fire control officer in the bridge is damaged. Do you instead have local fire control officers with each of the batteries in case central fire control is out of contact.

You could have fire control coming from a couple of redundant command centers, you know, just in case the bridge eats a nuclear warhead, or something. Failing that, one might assume that the programmers would be intelligent enough to design something known as a fail-safe. Something that would go "execute offensive/defensive fire plans as long as some command center is alive and on the network, but execute only defensive fire plans/shut off weapons should all command centers go offline."

Your sensors systems will probably handle most of the target analysis and selection. There is probably far more data than a single person can handle so do you have one person taking the computer's word for things or assign a small team to double check the input incase some new stealth technology from your oponent takes advantage of a flaw in your algorithms.

One might surmise that a civilization capable of building starships would be capable of building computers capable of doing this sort of pattern-matching themselves. Your "small team" is still going to get the data pre-digested by the computers, since a whole group of people drowning in data isn't going to produce an answer any faster than one person drowning in data. Yes, humans are fantastic pattern-matching systems, but this comes at the expense of an alarming false-positive rate.

You only have as much redundancy as the least redundant system if human input is required they are one of the systems. There is no point having a tripple redundant failsafe system that requires human input and only having 1 crew member and 1 station to control it.

Again, you would imagine that there'd be some redundancy in command and control aboard a starship. Of course, I feel the need to point out that in the case of minimal crew, these people would be placed in the safest part of the ship, like under a meter of armor and inside a central heavy water tank. If the one station in the one command center is destroyed, then chances are, the whole ship is fucked anyway.

[Enigma]

Uh where are you getting estimates for the Miranda from? A Miranda should be comparable to a Connie in terms of crew complement because they have similar internal volume. If a ship like the Defiant can have 50 dudes on it then the Miranda, which is a lot bigger, should have just a bit more.

Got them from Memory Alpha.

EDIT: The tech manual states that the crew complement was 220 but in TNG "Unnatural Selection" a U.S.S. Lantree was a Miranda Class supply ship with a crew of 26.

[Enigma]

Also according to ex-astris-scientia, the Brittain had a crew complement of 34.

AFAIK those low numbers are there because the ships were used in a non combat role such as a supply ship and therefore did not need extra personnel.

[Uraniun235]

As an example of this, consider the original Enterprise in The Search for Spock. The standard crew was something like 400, right? But Kirk and company were able to fly it, even badly damaged, with only a crew of five.

Didn't the automation freeze the first time they tried to do something more complicated then fly to point B, then completly overload the first time they took damage?

That automation system was also probably improvised in the field by Scotty, who would have designed the system without expectation of it having to cope with anything more challenging than getting from point to point, and which was attempting to run a ship that was never properly repaired after being mangled in combat. It's entirely possible that the automation system was itself relying on backup systems to function, with no redundancies left to switch to after taking damage.

The Enterprise-D, by contrast, is clearly designed to be highly automatic. We've seen Picard and Data fly the ship solo before, with Data even flying it solo while locking out the ship's systems from the rest of the crew. My guess is that you could probably program the Enterprise computer to operate sans crew in friendly space for awhile; fly here, pick up cargo, fly there, do a sensor scan, etc. Of course, you'd eventually need to have someone perform maintenance on the ship, and it wouldn't really be the best use of the then-most advanced starship in Starfleet, but it could be done.

[Stofsk]

Uh where are you getting estimates for the Miranda from? A Miranda should be comparable to a Connie in terms of crew complement because they have similar internal volume. If a ship like the Defiant can have 50 dudes on it then the Miranda, which is a lot bigger, should have just a bit more.

Got them from Memory Alpha.

EDIT: The tech manual states that the crew complement was 220 but in TNG "Unnatural Selection" a U.S.S. Lantree was a Miranda Class supply ship with a crew of 26.

I'd go with the tech manual. The two examples you gave were clearly supply ships and could get by with a small skeleton crew.

Even 220 seems low to me, but that could be why the Miranda became more common than the Connie. The latter in comparison had twice the crew, but not much more increased volume - and the Miranda seemed to be just as combat capable as the Connie was and we might be able to speculate that it was just as capable in other roles like surveys and exploration. But with half the crew complement, you could get more bang for your buck so to speak i.e. two Mirandas for the price of a single Connie. Hence, the reason why we keep seeing Mirandas decades later during TNG and DS9.

[Stofsk]

The Miranda crew couldn't have been too much bigger than Khan's gang. They ran it, and the Reliant's crew apparently was able to trade places with them on Ceti Alpha V too.

That's not conclusive though. Remember Khan's group also took over the Enterprise in 'Space Seed'. And yeah he did want to convert some of the Enterprise's crew, but that's because only Khan personally had read the technical manuals and the rest of his group had just been thawed out. He still needed experts to help run the Enterprise. I don't think it was a case that Khan's group didn't have the numbers needed to run the Enterprise.

[Simon_Jester]

I'd go with the tech manual. The two examples you gave were clearly supply ships and could get by with a small skeleton crew.

Even 220 seems low to me, but that could be why the Miranda became more common than the Connie. The latter in comparison had twice the crew, but not much more increased volume - and the Miranda seemed to be just as combat capable as the Connie was and we might be able to speculate that it was just as capable in other roles like surveys and exploration. But with half the crew complement, you could get more bang for your buck so to speak i.e. two Mirandas for the price of a single Connie. Hence, the reason why we keep seeing Mirandas decades later during TNG and DS9.

220 may be a 'limited crew' option, or maybe... consider this.

Suppose that the Mirandas have a basic crew responsible for maintaining the ship, plus... call them 'modular' groups who do different things. The ship might not actually carry enough people to respond to a medical emergency and do survey missions and land security troops to guard an isolated facility, the way a Constitution would. Instead, you'd load them up with the right selection of personnel at base.

Am I making sense?

[sirocco]

But seriously what would be the more expensive? A fully automated starship or a semi-modular one manned by an expert crew with the sufficient quantity of spare parts or the knowledge necessary to make some?