My question stems from What would a space fighter look like?

After reading through @dsollen's question can a human fighter pilot fully grasp 3 dimensional frictionless movement in space? I started questioning how the pilot would control the ship.

My interpretation of the ship is a sphere with multidirectional thrusters, and guns on multiple sides so that the pilot can shoot in any direction at any moment:

Ship design

My issue is that the pilot would struggle to do anything inside this ship if it were rotating (which is extremely likely in any situation). How would the ship be able to rotate the pilot to face the direction of travel, and how would the pilot be able to control their orientation manually?

Please note: The idea of a human-piloted fighter has already been established and this question is about them and not about whether they are in any way useful or effective.

  • 5
    $\begingroup$ To anyone arguing about pilot being obsolete: here is a question about how to make pilots viable, and I just started a bounty on it! So please, divert your energy to something constructive, OK? $\endgroup$
    – Mołot
    Aug 2, 2017 at 20:18
  • 1
    $\begingroup$ For a stories of scientifically accurate space combat, check out Larry Niven's stories set in Known Space, particularly Protector. One thing about space is it's so big, lasers would often take some time to reach their targets in a plausible space battle, so space "dogfights" would likely be slow and long. Nimble maneuverability and rapid aiming would not likely be useful. $\endgroup$ Aug 3, 2017 at 1:07
  • 1
    $\begingroup$ My advice is stop reading dsollen's questions or answers. Everyone knows that guy is a crackpot who insists on talking about himself in the third person! Also dsollen upvotes this as a very good question ;) $\endgroup$
    – dsollen
    Aug 3, 2017 at 11:53
  • 1
    $\begingroup$ This isn't a full answer, but you may want to read that crackpot dsollen's question here; worldbuilding.stackexchange.com/questions/17139/… Which will give options for mitigating the G force limitations of rotating and generally help a pilot survive flying this sort of craft. $\endgroup$
    – dsollen
    Aug 3, 2017 at 12:58
  • 1
    $\begingroup$ @depperm you are correct, however that is assuming that the fighter turns to face its enemy. To have minimal surface area exposed from all angles, a sphere is used. Many of the answers in this question assume that the pilot sees using screens, removing the need to turn and face the enemy and therefore the need to turn the fighter rapidly. $\endgroup$
    – Aric
    Aug 3, 2017 at 13:28

16 Answers 16


Okay, time for me to weigh in, a question like this practically invites me. Unfortunately this is difficult to answer without far more information about your world, and in particular what you have done to justify humans fighters instead of AI or missles being used.

For this answer I'm starting with a key assumption, high G manuvers are common. This seams likely in a close combat battle unless you go out of your way to prevent it. Much of this question is potentially invalidated if you remove high G meanuvers, either by making it too costly in terms of fuel, adding innertial dampeners, or some more complex handwave such as my claim that shields worked best with low acceleration thrusters.

This answer also presumes a number of thrusters on a ship. If you limit the ship to only say 3 thrusters or had a primary thruster that was stronger then the others (both plausible for cost reasons, it may be cheaper to make fewer thrusters or lowe output thrusters an demend economical to trade off less meanuverable crafts for the ability to make more) then this answer may be invalidated. If the number of thrusters is limitd enough the control scheme could change, but this answer still is quite likely. If there is a single thruster with more output then others then that defines your 'forward' and this could redefine much of the answer.

Where should the pilot face

the pilot would presumably would reside within a control hub orb at the center of the craft. This orb rotates so the pilot is best positioned to tolerates the G forces associated with acceleration (see Low-tech inertial dampener options ) This rotation is automatic and the pilot doesn't control it. The primary motivation is simply to allow the pilot to survive higher G forces by positioning his body in the most optimal position to endure the force.

In fact to prevent the pilot from crushing himself with G forces, while still allowing him the ability to make use of all acceleration he can handle, the craft would likely have some sophisticated soft (non-intellegent, just well programmed) AI that considers the pilots position before deciding how large a thrust to exert. The pilot will not be allowed to achieve 'maximum acceleration' when his body is in the wrong position because max acceleration would kill him. Instead he can only accelerate up to some acceptable G force, and then as his body rotates into a better position the maximum acceleration allowed to achieved in that direction will be increased since the computer determines the pilot can now survive that level of G force.

This is all done purely to protect the pilot, but as a side effect the pilot now has a very clear orientation. The direction(s) he is currently accelerating towards are 'forward' to the pilot. Keep in mind that the pilot will sense his 'forward' as being the combination of all forces on him, he may have three fruster firing at different directions & rates to propell him, but his sense of 'forward' will be the vectoral sum of the acceleration created by those thrusters.

In the case that the pilot is not accelerating at all then 'forward' will be defined as the last direction the pilot was accelerating towards. The pilot may have an option to override his 'default orientation' in situations of no (or very little) acceleration for his own covenience, such that he is looking the direction that he is most interested in looking, but during the heat of combat when he will constantly be accelerating in differnt directions he will always be using his acceleration direction as his default 'forward'. From a programing perspective likely he can set his default forward if acceleration is uner 1 G but the moment he experiences more then 1 G worth of acceleration the computer will immediately revert him to 'acceleration=forward' mode which is the prefered mode for heavy movment or combat.

Eyes in the back of my head (and top, and bottom...)

Now that we have a forward, how will it be displayed to the pilot? Most likely he will have a 180 x 180 (so up & down as well as 2 D plane) degree display, either actuall screens on the wall or more likely some sort of headgear that displays the screens infront of his eyes. He will likely have his screens split such that half his screens are 'forward facing' and half are 'aft facing', so he can see all degrees of the sphere by changing which screen he looks at. Actually the screens would likely be configurable and allow him to do things such as making his aft or fore screen view larger or break up more screens depending on his needs, one sign of a good pilot would be being able to modify his screens to give him the most 'optimal' view for him and each pilot may have a different prefered display, in much the same way that people playin MMOs may have widly different screens as they install and remove addons to configure their display to display information in a way they prefer. The point though is that he will have the overlay to display everything.

This overlay will constantly adjust his view as he is rotated due to changes in acceleration, to keep the center of his fore display facing the 'forward' direction (ie the way he is accelerating).

Actually, there is a slight translation time here. He can't be instentaniously rotated towards a new point acceleration (there is some time from when he starts firing his port thrusters and when he is rotated within his cockpit so his port frusters are now his aft thrusters). It would also be confusing to have the screen transition immediately to show a different point of view as 'forward' to the pilot, he needs to watch the screen rotate to a new 'forward' direction so he knows where something relevant (like an fighter craft he is chasing) is relative to his new 'forward'. These two issues work nicely together though.

Presumably when he starts to accelerate in a new direction both his physical body and the screens sense of 'forward' will be rotated at the same time. He will see the screen rotating as he 'feels' the rotation, this will help to keep his orientation during the transition between the 'old' forward and the 'new' forward. While he is rotating his tactical display will likely depict an icon (and possible an 'arrow' along with it) that will show him where his new forward is (ie where he is rotating to face) with the icon rotating towards the center of his vision as he is adjusted towards the new forward.

The Enemy's gate is wherever it happens to be

Notice this makes direction entirely relative, there is no absolute direction, I can't tell you that the enemy gate is down with this approach. The enemy gate was fore a minute ago, now it's aft, sone it's port/fore/up etc. I don't consider this a problem. You don't really care about an aboslute position, all you care about is where you are relative to key tactical points and/or threats.

If were assuming a hectic close range space battle (again, probably not realistic in real life, but implied to be the case in your world) then things are changing all the time. Enemy crafts are moving around you, they are never at the same place twice. Unlike Ender's game there was no final destination you're headed to and that your done once you reach. You may think of your enemy capital ships as 'down', but your fight isn't over just because you reach them, your just fly past them and rotate around for another pass at them after all. Nothing is really 'down'. All you care about is where you are relative to the things you want to be relative to.

This information is better served with a good tactical display! You already are being provided some semibalance of a 360 degree display, so if you want to know where a gigantic capital ship is just look, it should take a split second to identify it in your display. To make it even faster (because in combat split seconds matter) your have more complex targeting and tactical displays overlaying your 360 HUD. You could arrange for a little colored arrow to be projected from the 'center' of your screen in the direction of key locations you care about, like your home ship and the enemy capital ship your attempting to bomb, so you know what direction to look to find them. You can toggle targets to have them lit up on your screen and similar color coded lines to show where they are. A tactial display showing you what direction your relevant points of interest are is more useful. In a sense you can have multuple 'enemy gates', ie points of tactical importance you want to always be aware of the location of, at one time.

The enemy gate is also at starpoint 125x234x3

Okay, so there likely is also an absolute direction that is predefined by every fighter's computers. This can be used for a pilot to communicate concepts such as "fighter inboud as 6 oclock " in a method that has meaning to other pilots. This coordinate system would likely be mostly invisible to the pilots though, instead of trying to figure out actual cooridnates they would ping some location on their map and their com system would relay the location to other's maps and simply know where to ping relative to the pilots sense of forward. Alternatively some third party would be feeding updated tactical information. So yes there is a way to give absolute direction, but it's not something a pilot needs to think about or consider in the heat of combat.

that's one funny looking steering wheel

I haven't said it explicitly, but al controlls would transition at the same rate that the pilot own vision was rotating. If he pushes the fore control his thrusters will push to move him in the direction he is currently 'facing' in his display, whatever that direction is. As his display rotates towards a new 'forward' his controlls will adjust their definition of what 'fore' is accordingly.

So how does the pilot use this display to steer? That's a more complicated question. Other's have suggested joysticks, trackbars, etc, they won't help. All the control schemes we currently use as human only really work for two demetions, not three. The 3 dimentional control scheme I can think of would litterally measure a hand's motion in 3 dimentions, but doesn't work well as soon as a force can cause the hand to move when not intended (ie, when you suddenly have 4 G of force pushing you down because you had to duck under an incoming missle and you can't suspend your hand as easily under that force).

As such I don't think a single control scheme can be used. I think we need two control schemes, utilizing your equivlent of a joystick for controlling two dimentions, but having two of them to combine to allow control of all 3 dimentions at once. To put it more simply, one joystick for each hand. So which directon does each joystick control?

I spoke about our sense of acceleration as being a single senes that tells us exactly what direction were traveling...but I lied. We actually have effectively two senses for acceleration, our sense of up & down is rather different then the others. In particular our sense of 'down' is far more accute, due to the major plummitting sensation in our stomach we get the moment we have more then 1 G of force facing down. As such if we have a two controll scheme it makes sense to have those controlls be one for the standard 2 d plane and a second for 'up' and 'down' plane, combining to allow controlling all four planes.

Were facing war, you must respect the gavity of this situation

full 3 D space flight, particularly with such heavy accelerations as I anticipate in space, suggest a new wrinkle withour controll systems. At any time you may be experiencing multuple G worth of force pushing in any direction. This means that no matter how you lay out your controlls it's possible that your pilot may have heavy G forces pushing the pilot's hand 'down', ie making the pilot push his controlls in a direction he doesn't want to because he can't hold his hand steady agains the G force he faces.

This situation doesn't happen in a fighter jet as much, but the pilot only risks facing high G in one direction at a time, and the controls can be adjusted accordingly. facing high G from any number of different directions at once is much harder to handle.

At the very least the pilot's craft will have to adjust partially for this. If the pilot is likely to face heavy G pushing his hand in a direction decrease the sensitivity of the controller (or even have the controlls actually 'push back' against his hand harder) by an according amount so that his pushing harder on the controller does not actually cause the controller to register that as a request to move faster in the stated direction.

Some other approaches, such as having the two different joysticks reside on perpendicular plans from each other, could also be used to address this, but I won't get into the subject much more then to suggest the possibility.

Your pulling me in multuple directions

One last hurdle to driving is the competing need to move in different directions at different accelerations. Remember, the pilot can only accept G forces up to a certain degree before they render him unable to fly, and, more importantly, the degree of G force he can tolerate is dependent on orientation of the force. He can survive g forces directed 'down' much better the ones directed 'up' for example. For this reason I suggested that the ship control the maximum g force allowed based of off the direction the pilot was facing.

This brings up an intresting question for how we define what it means to push our joystick as far forward as it can, ie we say we want to accelerate as fast as possible. Since 'as fast as possible' is different depending on direction were headed it can have odd implications. Imagine I push my joystick as far as it would go in a 45 degree angle, intuitively saying "I want to go as fast as possible in the fore and starbord directions" I can survive acceleration in the fore direction that are much higher then the starboard. Do I accelerate as fast as I can handle in each direction, meaning I don't accelerate diagonally as I might assume, but instead in a direciton that is far more fore then starbord? or do I limit my maximum acceleration fore so that I'm accelerating both fore and starbord at the same rate? The former seems confusing, the latter could get me killed if I'm trying to dodge a missle and my computer won't let me move as fast as I theoretically could to avoid it.

I believe thta the controlls always work like the later example, accelerating at the maximum allowed rate for each direction even if that means I accelerate faster in one direction then the other. However, a pedal (or other control) will likely exist to allow me to instead 'correlate' the directions, so that my acceleration is based off of the maximum acceleration in whichever direction I can least tolerate acceleration in. So basically the can toggle between most acceleration and more 'controllable' acceleration. If this were a pedel they would even get a degree of control between 'completely correlate these two directions' and 'don't correlate at all'. Though you could argue the inverse, where directions are usually correlated and a pedal allows maximum acceleration in any direction instead, as being the default. The key point is the likely can alter the control schemes.

I tend to put on weight when stressed

Again the computer is trusted to decide 'maximum allowable acceleration', but what is that? Humans can high g forces for short term that they can't survive over a long term. In addition a human who has been experiencing a high sustained G force for awhile may be stressed and less capable of surviving a suddent short term higher G force in another direction, even if they could have survived that under other more ideal situations, and of course some pilots can simply endure higher threasholds then others.

If we assume thrust is cheap (which it is if your having highly meanuverable crafts at all....) then the craft will likely be designed to allow thrust up to the max a human can survive even for short times, to allow more meanuvability. But the computer may not allow someone to try to travel at that for sustained periods.

I imagine there would be a(nother?) pedal that is the "Oh $%!#" pedal, this is the one that says the pilot needs to accelerate now no matter what to get out of the way of something. This pedal would loosen the definition of 'maximum' acceleration to allow higher G forces to be experienced by the pilot. When the pilot is willing to risk temporary high g to do a meanuvour he presses this pedal to allow it, but the pilot can't accidentally go so fast that he knocks himself unconcious unless he has pressed this pedal, it's the laxing of his own safety procedures when he judges it necessary. This is definately a pedal to allow more nuainced control of just how much 'extra' G force to allow, a slight press on the pedal may simply mean "I want to go at a higher sustained G then usual, but nothing dangerous" while a full press on the pedal says "I'm willing to pass out half a second from now because that's the only way to avoid coliding with a friendly fighter and killing both us.

This is war, man up!

Now an aside, the most realistic is that your fighter craft will have at least two men in them, much like modern fighters. This frees your pilot up to use him hands entirely for steering, no taking his hands off the controls for anything else.

This is relevant becuase my controll scheme has the pilots hands busy at all times. While this scheme can work with a single pilot it works far better if a second person was in the fighter handling controlls that the pilot can't afford to worry about while focusing all his energy on driving. Driving will be harder here, requiring paying attention to multuple angles and fighting the G forces to keep your hands on the controllers right etc, so the pilot won't be able to afford as much attention to other activities as moder fighter jet pilots can either.

A sophesticated AI, especially if combined with audio input to controll less-critical systems, could likewise help here if you don't want a second pilot.

If you have a single pilot expect the 'joysticks' the pilot uses to be very complex with many controls built around the joystic to be toggleable with button presses. Though only so many things can be on a joystick, some controlls must be placed moer out of the way to avoid accidental activation.

Shooting down some suggestions

Now lets address the related question of aiming the guns, which has been brought up We are already more then capable of writing a soft Ai capable of aiming a turret gun at a moving target with more accuracy then a human pilot ever could. Furthermore, it would take too much focus for a pilot to focus on both aiming and moving, one act would distract from the other. As such the logical option is that a soft (Ie, not intellegent, just well programmed) AI would be responsible for fireing. This has the added advantage of allowing firing at multuple targets at once, making full use of every turret despite the pilot only being able to look at one at a time.

Unfortunately it's also boringly practical, it doesn't feel like an interesting space battle if your only job is to drive and trust your computer to kill everything. It also makes your pilot feel rather worthless and just reiterates the idea that humans don't belong in space fighters, which is something you've clearly implied you want to make happen.

One option is to say that there are only so many guns on the sphere, that they can only cover certain angles of fire, and possibly that they have a slow recharge time before being ready to fire again. This places more tactical challenge on the pilot. He won't be aiming exaclty, but his job is to make sure an enemy is within range of the turrets that are ready to fire.

This would give a feel much like navel battle where one didn't so much 'aim' their cannons as they focused on trying to 'cross the T' to arrange for their cannons to be facing in the general direction of the enemy while the enemy wasn't facing them and trust the cannons to fire. The tactics are all about knowing where you want to position yourself to ensure a shot can be fired, and even more so knowing where your enemy wants you and not being in those spots. A skill pilot may even notice that the enemy just fired from his aft cannon and that there will be a X second recharge before that cannon is ready to fire again and thus he can move in for an easy kill while one of the gunns is unable to cover the pilots flank etc.

Alternatively you could add a second man in charge of acting as a gunner (there is no way the pilot can fly well and shoot well in this scenario, not unless you greatly limit both number of guns and thrusters). If so they gunner will likely have a setup similar to the one for the pilot, except that in addition to the usual 360 display he has the ability to toggle which 'gun' he wishes to control and he gets a zoomed in display with that gun but...well it really takes allot of work to justify having humans gunning and the solution for how/why the humans are aiming in this case depends largely on what handwave excuse you used to justify guns being mained by a pilot.

  • $\begingroup$ "what you have done to justify humans fighters instead of AI or missles being used." Imagine that cyber warfare is so advanced that any computer system using a network is immediately dangerous and hackable. Therefore, isolated computer systems within ships, with no connectivity to any other system, are required. $\endgroup$
    – Aric
    Aug 3, 2017 at 19:54
  • 2
    $\begingroup$ Also, +1 for being the most magnificent answer i've seen in my year on this site :D $\endgroup$
    – Aric
    Aug 3, 2017 at 20:00

The enemy's gate is down

While we know that in practice it would all be drones and allow no manual control, a story needs a hero and pilots make good heroes.

However as anyone who's played asteroid will tell you, even in two dimensions controlling flight with 0 resistance is a fairly difficult thing to do. That's where pilot aids come in.

Let the AI/autopilot or other similar system handle stability control. A feedback system from the joystick that maintains rotation while there's pressure and stops it when the joystick is released. Gryroscopes can track the actual angles involved, that's a very basic technology, the computer fires the thrusters to stabilise the movement. The pilot just needs to point and shoot.

This of course requires a switch to disable it for plot reasons. You'd be mad to disable it in practice, but the hero can cope, otherwise he's no hero.

There are multiple options here, you can have it accept lateral movement, correct lateral movement, accept or correct lateral movement only on command. All of these have advantages in a theoretical fight. Spin/tumble should always be corrected unless the pilot disables it.

Additional notes

More thrusters fewer weapons

6 thrusters only handles lateral, you need off axis thrusters to handle rotation. 6 clusters of 4 or 5 thrusters will do it.

If the ship has weapons in all directions, there's no need for the pilot to handle complex maneuvers. A bit of hand waving to limit his arc of fire to something along the primary axis of his ship.

  • $\begingroup$ The autopilot sounds like Flight Assist in Elite: Dangerous. So, you're suggesting a setup where the pilot can control forward speed alongside pitch/roll/yaw and the AI will use the thrusters to make it move like a plane? Also, the AI could recognise any external rotation which does not match the pilot's inputs and counter it (for example, correcting rotation while spinning after a collision). $\endgroup$
    – Aric
    Aug 2, 2017 at 13:07
  • $\begingroup$ @AricFowler, the pilot will want lateral/vertical movement as well, strafing alongside a capital ship for example. $\endgroup$
    – Separatrix
    Aug 2, 2017 at 13:11
  • $\begingroup$ of course, so in that case two joysticks (rotation and x/y thrusting) and foot pedals for z-direction movement. $\endgroup$
    – Aric
    Aug 2, 2017 at 13:12
  • 1
    $\begingroup$ @AricFowler, go scifi! A sphere suspended in a control field allowing and sensing full movement with suitable feedback. The distance it moves off centre proportional to the available percentage of thrust. $\endgroup$
    – Separatrix
    Aug 2, 2017 at 13:15
  • $\begingroup$ You can only use drones if you can communicate with in something very close to realtime and over the distances of space combat that means Faster Than Light otherwise the signal lag will get you every time, in practice you need an onboard pilot either AI or Human and AIs aren't that dumb. Even if you have Faster Than Light control systems those are vulnerable to electronic warfare if the enemy knows your frequencies anyway, no drone starfighters. $\endgroup$
    – Ash
    Aug 2, 2017 at 15:07

Have the pilot/gunner in a 3 dimensional rotating turret sphere that he could rotate to any orientation independent of the ships orientation with the weapon systems tracking the direction the pilot/gunner is facing. This could also counteract rotation of the ship (especially if the crew pod is centered on the crafts axis of rotation/ center of gravity) to keep the pilot/gunner facing in any desired direction.

The best example in fiction I can think of is the gunner seat in the Gunstar from The Last Starfighter movie. The pilot was oriented forward in the ship and flew the ship, while the gunners seat rotated to allow the gunner to point in any direction, and the ships guns tracked the gunners orientation and fired where he pointed.

Here is a fine paint drawing showing the concept. enter image description here

  • 1
    $\begingroup$ This didn't work too well with aircraft (turreted fighters were popular in World War I, but had almost entirely vanished by World War II), so why would it work any better with spacecraft? $\endgroup$
    – Mark
    Aug 2, 2017 at 21:57
  • $\begingroup$ @Mark aircraft guns were mostly replaced with missiles (i.e. unmanned drones), which is likely what actual space combat would look like (which is boring for storytelling), but for the fighter design in the question, a manned unidirectional space fighter required to fire in multiple directions, it is a very usable option; as shown by its previous use in aircraft. $\endgroup$
    – Josh King
    Aug 2, 2017 at 22:08
  • $\begingroup$ This is seriously cool, and can fit many styles of space opera from gritty to clean. The actual use of similar technology in ww2 aircraft ensures it's somewhat familiar to the audience. $\endgroup$
    – SáT
    Aug 3, 2017 at 9:59
  • $\begingroup$ I understand the idea, but having the gunner actually rotate seems pointless. you can project whatever you want on the gunner's screen, he can 'rotate' his viewpoint by changing what's projected on a screen without actually rotating. He should rotate, but his seet should rotate so he is always facing forward (ie towards the primary direction of acceleration) to lower G forces he experiences regardless of where he is pointing his guns. Always face him in one direction, but let him change the area displayed to him that his guns will be fireing. $\endgroup$
    – dsollen
    Aug 3, 2017 at 15:47
  • $\begingroup$ ....furthermore, this presumes he has a gun he can rotate in any direction. This is not the same as what the OP postulated. He will have some X number of guns at specific points on his ship. Those guns likely will be able to change the angle they fire at, but only so far. with set point defenses buit into the spere there will be a blind spot. It would be more like him 'picking' which gun to aim then picking where to fire thta gun, depicting his view as 'moving' a single gun is inaccurate. It also can be slow, if he needs to fire 180% from his current angle he should be able to do it fast $\endgroup$
    – dsollen
    Aug 3, 2017 at 15:50

I wouldn't worry about that operation. You only turn in the direction you are flying as a convenience. There's a wonderful scene about this in the book version of Apollo 13. They explain how they're going to do a burn to turn the spacecraft around to face forward. In the book, they explain that this is a completely wasted maneuver. They get nothing out of the spent delta-V. The reason for turning around to face forward: there's an army of engineers with slide rules on the ground making sure this flight goes smoothly, and they were worried that one of those engineers would get the sign wrong on an equation if they weren't facing in what terretrial-bound humans think of is "forward." It was decided that the wasted fuel was worth avoiding the risk of a human error in the slide-rule era.

The standard way to manage orientation is to have clusters of small attitude control thrusters. They don't have to be big. Changing the attitude of a ship is way easier than accelerating the craft in a meaningful way. You could also spin up gyroscopes to save fuel on small maneuvers like the ISS does. They have their limitations, but they are much more flexible than bleeding precious mass out into the void.

I'd give them a trackball control -- you only have 2 axes that matter for the purposes of aligning guns. Because the roll axis isn't as important, you could probably actually set it up so that the ball spins inside the socket to control roll.* On their computer screen, you could have a visual like Google Sky showing the constellations as reference. I'd probably have two modes. Auto mode, where they point in an orientation and click a button to say "please orient me this way as efficiently as possible," and a Manual mode which is never used for any reason, but your hero will probably demand a manual mode so that they can override the computer. Heroes are that way.

* I've never seen an arcade game that tracks the ball spinning in the socket, but now I'm a little curious if there was one!

  • $\begingroup$ +1 for sticking with it and not questioning the question. + references to real world, actual, information. $\endgroup$
    – AnoE
    Aug 2, 2017 at 22:08

I will not try to optimize the 6 you found for the number of thrusters or your concept. Somebody build that ship, some very poor guy had to fly it into battle and it apparently started rotating sometimes because something broke perhaps. I will also not say that one needs a window to look out of, one should not ask too many questions about science fantasy concepts. A computer screen is necessary if only to help see - space is too large for human eyes. I will call the ship the "death trap" for historic reasons.

So the answer I think is rather simple: You need a computer program that rotates the pilot if needed. When would it be needed? Humans are not that great at being accelerated in any direction really (another reason why I wouldn't use humans for this), but we are best if it is in the direction we are facing. So if one accelerates quickly, you either press a button or the autopilot does it. A program could easily identify the need for rotation and start spinning the "chair" or "cabin" or whatever so the pilot faces in the direction of acceleration. I do not think it is so much an issue of "how can he/she still stay in control" and more one of surviving acceleration. So during those periods, you need at least computer assisted steering. This is where (genetic) enhancements come into play - improved humans could maybe survive more acceleration.

The other case would be if there was some kind of malfunction or hit or whatever that caused rotation. But as I see it, this would be the exception and the autopilot would easily correct this again with the thrusters. If this happens to be common, you can have your piloting cabin float in some kind of medium - so if the outside starts to rotate, the inside does it much slower, allowing time to counteract.

When will it not be needed? Any other time. There is no need to constantly rotate the pilot to constantly face in whatever direction is defined as forward.

So the computer will also counteract any kind of rotation that has happened for some reason. If both the thrusters and the rotating cabin malfunction, well, another reason why I wouldn't put a living human into that "death trap". Btw, think 2 sides fighting a war, one uses human pilots, the other one doesn't. The 2nd faction will have much faster accelerating spaceships and gain an instant advantage.

For manual controlling anything: Do not think so literal, we are dealing with advanced technology here. Even if there is no rotating, this is almost impossible to do. The pilot would see everything he needs via computer screens and can quite easily select a path, maybe even via touchscreen. Doing a calculation, especially with 6 thrusters, to find the right linear combination to do anything manually - now I'm starting to understand why the fighter keeps on spinning. My car already does most of the thinking for me, the same would be true for an advanced space fighter. It is very easy to calculate for a computer how the thrusters should be powered, but very hard for a human at the speeds required.

At the end, for the human it will almost look like a computer game or simulation if you will - he will have some input devices, maybe something fancy like mind control, maybe a good old joystick, maybe something very similar to what fighters have right now , and a computer program will translate that input for him.

I once saw a demonstration of a medical instrument that allows for 3d movements and rotations of equipment (or something) during operations, I do not remember the name though. Something like this would also be possible here. If you do know the name, please let me know via a comment.

Many modern cars one can simply buy already do most of the thinking, it would be insane to assume that spaceships have less advanced electronics. This is why I commented on cyborg and mind reading devices. It is really not that far fetched compared to space death traps. If you can directly input your thoughts, you are much more maneuverable in battle. People are already working on such devices. But once again, keep also in mind that self driving cars are already a reality.

So, in short: If you absolutely have to put people into fighter spaceships and then ask questions about how this is possible, use enhanced super-humans and pretty straight forward solutions to rotation

  • $\begingroup$ With regards to the window, i was going to get to that after this question is answered $\endgroup$
    – Aric
    Aug 2, 2017 at 13:19
  • $\begingroup$ If you feel there needs to be a window to look out of, please state so in the question. That would completely change the nature of the question. I just got the impression we are talking about sci fantasy type fighters that usually have a window because it looks good in a movie - but do not ask questions then please $\endgroup$
    – Raditz_35
    Aug 2, 2017 at 13:34
  • $\begingroup$ What I meant by the first comment is I have not specified how the pilot will be able to see, and it's up to the person writing the answer to either ignore it completely or come up with their own solution. $\endgroup$
    – Aric
    Aug 2, 2017 at 13:42
  • $\begingroup$ @Raditz_35: yeah, no worries. Cleaning up my comments and the downvote btw as the answer is fine with me now. Thank you! $\endgroup$
    – AnoE
    Aug 3, 2017 at 10:07

In the movie "Apollo 13", the crew has to make a manual burn to adjust their orbit while the ship's computer is shut down. The crew (with some help from their flight controllers in Houston) aim at the terminator line on the Earth and use it as a reference point while firing the engine. While extremely dramatized in the movie (the real Apollo 13 did not bounce around like a pinball), this illustrates that manual orientation and navigation in space is indeed possible.

Unless you are using some sort of magical drive, real spaceflight is actually rather "slow and majestic". Distances are vast, and you make deliberate movements with your spacecraft in order to minimize the use of fuel and prevent things like over rotation or spinning out of control. Even in Earth orbit, you might have to take several orbits to close to firing range with your space fighter, so you will have lots of time to make adjustments, line up your shot and so on. (Of course your opponent is also using his time wisely.....).

The reality for near term and plausible mid future "space fighters" will probably resemble the way the former Soviet Union handled things. Long range radars track aircraft and ground controllers (or controllers on "spacecraft carriers") will be sending detailed instructions to the pilots to vector them onto the target. If possible, the controller will even set up shots for "beyond visual range" missile shots. The pilot is only released to engage once he is in visual range of the enemy, in which case the controller can't help anymore. The pilot uses his thruster clusters to point the nose or missile launchers at the enemy (probably based on a combination of visual and radar cues) and takes the shot. One the enemy is destroyed or the fuel warning comes on, the flight controller then provides instructions on where to point the spacecraft, what burn to make and guides the craft home.

  • 1
    $\begingroup$ It would be interesting to write a story about combat where the limitations are not ammunition or fuel, but food and oxygen to survive the extreme number of orbits needed to make contact with the enemy! $\endgroup$
    – Cort Ammon
    Aug 2, 2017 at 14:44
  • $\begingroup$ Indeed! Look at youtube videos of Apollo spacecraft docking with the LEM to get an idea of the sorts of time involved for even simple manoeuvres with a cooperative spacecraft. A space dogfight between a militarized Apollo and Soyuz would probably resemble a naval battle during the age of sail more than anything else, and making the opponent break contact due to lack of supplies might well be a tactic. Apollo is much larger and more capable, so they can wait out the Russians if needed. $\endgroup$
    – Thucydides
    Aug 2, 2017 at 19:12
  • $\begingroup$ But then, they did not have Flight Assist in Apollo capsules... in fact, I guess every simple pocket calculater today has more calculation power than the total of CPUs on Apollo combined... $\endgroup$
    – AnoE
    Aug 2, 2017 at 22:10

One way to do it would be to arbitrarily establish "down". Our brains are evolved to think that way.

The ship would fix on some distant bright object and set that as down. Our inner ears do a lot to establish which way is up and which is down (the eyes help too). The inner ear works via inertia: movement of fluids in the semicircular canal and weight of otoliths. image from http://what-when-how.com/wp-content/uploads/2012/04/tmp15F75.jpg enter image description here


Sensory information about motion, equilibrium, and spatial orientation is provided by the vestibular apparatus, which in each ear includes the utricle, saccule, and three semicircular canals. The utricle and saccule detect gravity (information in a vertical orientation) and linear movement. The semicircular canals, which detect rotational movement, are located at right angles to each other and are filled with a fluid called endolymph. When the head rotates in the direction sensed by a particular canal, the endolymphatic fluid within it lags behind because of inertia, and exerts pressure against the canal’s sensory receptor. text from http://vestibular.org/understanding-vestibular-disorder/human-balance-system#

If these fluids and the otoliths were replaced by magnetic items of similar density, the pilot could operate normally during shore leave. In the ship, the ship could make a magnetic field corresponding with the predetermined arbitrary "down". It is not artificial gravity but the pilot would feel "down" and could orient himself.

I could imagine the ship swinging around to facilitate operations in a head up, tail down orientation for the pilot. If you are going to rely on human biology to do a job, it is understandable to make a few accommodations.

ADDENDUM I see with @Aric Fowlers comment below I was unclear. I address this question from OP: /My issue is that the pilot would struggle to do anything inside this ship if it were rotating/ I interpreted "struggle" to mean that the sensation of rotation is disorienting. Which is definitely true; rotation at varying speeds around varying axes sounds like the most nauseating type of carnival ride. My answer is to make it feel to the pilot that he is not rotating. Whether looking out a window or watching a screen it will feel as though he is sitting at a desk on land.

If the question is just the physics of how to rotate a sphere using "6 perpendicular thrusters" then the answer is you cannot. Draw the vectors and you will see why. One must have a tangential force to produce rotation. Your thrusters must come off of the perpendicular. If a thruster can bend at 90 degrees it can produce a tangential force and cause rotation. You could give 1 thruster 180 degrees of freedom and it could produce all rotation. You could give each thruster the ability to bend 90 degrees in 1 direction and a combination of thrusters can produce any rotation.

With the structure of the ship I can think of no reason to rotate on purpose, unless the pilot is peering out of a glass window. Even my 2008 minivan has a backup camera. The ship can move any direction and shoot any direction. It will have screens showing those directions. It is symmetrical; there is not really a front and back. Front and back is a relic of vessels that have a front and back.

That said, the ability to rotate the ship would be good because it might start rotating from an external applied force and you will want to stop that if the ship ever docks or lands. Otherwise it will act like a pingpong ball with spin on it and skitter off into the hanger once it makes contact.

  • $\begingroup$ This doesn't directly answer the question: "How would the ship be able to rotate the pilot to face the direction of travel, and how would the pilot be able to control their orientation manually?" I'm asking how you would acheive "the ship swinging around to facilitate operations in a head up, tail down orientation for the pilot", not how to trick the pilot into feeling gravity $\endgroup$
    – Aric
    Aug 2, 2017 at 12:31
  • $\begingroup$ I've simplified it to multidirectional thrusters then. Thanks! $\endgroup$
    – Aric
    Aug 2, 2017 at 15:15

I'd like to suggest a different approach, one which plays to the strengths of the human sensorium rather than fighting against our basic nature; gun systems should all fire in a single direction as dictated by the pilot's orientation thus creating an arbitrary "forward" to the ball ship, this removes 1. about ten years worth of training for your pilots by allowing them to pick an arbitrary external orientation set up that works for them (this orientation scheme can be set up fleetwide, and in fact it should be for ease of communications and maneuver) without requiring true three dimension, 360 degree awareness and 2. removes a whole mess of aggravation concerning weapon systems redundancy, duplication, and power/ammunition access.

I can't take create for this solution it comes straight off the pages of the Lost Fleet series by Jack Campbell.


The pilot would set their orientation either to an origin point, a dynamic point, or an plane.

And then from there they would probably constantly reset it.

What this means precisely is imagine the outer ship, with all the guns/thrusters (let's only really talk about thrusters) on the outsider. You can imagine this has a sphere for now, though likely you would have some stronger thrusters and weaker thrusters, which you rotate around in order to accelerate in the correct direction. Some of these thrusters are orthogonal to the outer ship sphere, and some lie tangent to the sphere for the purpose of rotating the ship (these tangent thrusters are unnecessary if you imagine the ship as having an infinite number of thrusters at all points on the surface of the sphere). A more advanced ship may have thrusters that are at an angle between orthogonal and tangent, but that's confusing and just accomplish a mix of what the other thrusters do.

Now, the pilot cockpit, is another concentric sphere, which always orients itself towards the target. Imagine computerized ball bearings between the outer ship and the cockpit, which as the ship spins or displaces, these ball bearings all adjust the cockpit such that it is looking towards whatever the target the pilot set it on.

Re-hashing what the target would be, it being the objective makes alot of sense. This could either being another ship, docking thing, a race track that updates to a new point along a arc as you advance.

Another possible target could be a specific point in space that the pilot sets. Since everything is relative, they hit a button, and maybe their current position becomes the target, and now as they fly around, they always look at it.

The last possibility I mentioned, would be setting it to an axis, like defining their current facing direction as looking at the x y plane, and then regardless of how they translate their ship, the cockpit continues to face the x y plane.

This then requires the a joystick of course that allows the pilot to easily control their position relative to the target. I like to imagine this as a floating ball in front of them (for simplicity, I'm sure there is a better control system), where they displace it in 3d space to control their thrust relative to target.

From there, if we have our sphere with infinity thruster density across it's surface, it's easy to imagine. If it is a mix of stronger and weaker displacement thrusters in different locations, with rotational thrusters, a computer quickly handles the orientation of the ship using the rotational thrusters in order to support the pilot's thrust decision.

But in reality we wouldn't use pilots probably, just super sophisticated missiles

  • $\begingroup$ The answer was great right to the last sentence. I think it would gain if you would just remove that one. $\endgroup$
    – AnoE
    Aug 2, 2017 at 22:12

In my imagination, I'd always imagined the best controller to be a little model of the ship, that you can pull around by hand, giving you the full 6 degrees of freedom.

The closest to this that I ever found, and the best UI controller I've ever found for universal-degrees-of-freedom gaming such as Descent, is the SpaceOrb 360, which I believe is sadly no longer produced:

SpaceOrb 360

Unfortunately, it's a serial device, and the drivers aren't compatible with anything even if you manage to get one and a serial port adapter.

[If anyone's super interested in this, https://www.hanselman.com/blog/TheBestControllerForFPSASpaceTecSpaceOrb360ControllerWorkingWithWindows7UsingArduinoAndOrbShield.aspx has a great writeup about getting it working with newer hardware... while the OrbShield adapter is now called the orbotron, available at http://www.thingotron.com/ maybe (been a while since that site was updated)]

As for which position the pilot should be oriented, I'd say the answer is "facing forwards". In a sphere, that's going to be the current (or, most recent) direction of thrust. A simple way to do this is to have them in a weighted, dampened gymbal, like a ship's compass: no need for computer control, then.

Some computer control would be nice, so you wouldn't flip around when braking on approach to other objects; I'd say just locking the gymbal once you get close to other objects.

Another option which would be intuitive to the pilot would be to allow the pilot to just specify an object (earth, sun, space station) that their velocity is to be measured relative to, and then face them "forward" in the direction of that velocity.

Automated gunning will be important. You don't need to have a gunner following the gun, just tap on a thing on your scanner to lock onto it and shoot it. A manual gunner would never get the relative math correct (yes, this makes for a less dramatic story, I'm sorry).

A number of other things should be handled by the ship, too: approach, docking and undocking with other things, avoiding running into its its own projectiles, and plotting trajectories and thrust directions. You tap in your destination, and the ship calculates the best path through the system to minimize time, or fuel, or whatever.

But that's realism. For drama, the pilot should probably be involved in some of this stuff at a lower level than simply commanding.

  • $\begingroup$ That controller looks pretty cool, I'm sad it's no longer produced. Fortunately there are modern 3D mice offering 6 DOF designed for CAD, modelling, etc: 3dconnexion.com/products/spacemouse.html $\endgroup$
    – talrnu
    Aug 3, 2017 at 5:19

Here's my take on it. It's largely an argument against the need to orient the pilot of a space fighter, but I still feel it's a valid response to your question.

Firstly, my assumptions:

Single Pilot - To preserve the ideal of a fighter pilot in single combat, I've opted for a model where the spacecraft has a single crewperson.

Direct Manual Control - Assuming that the same restrictions that preclude computerised weaponry also preclude the use of computerised attitude control and so forth, we're left with direct manual control.

Orbital mechanics - Without a propellant several orders of magnitude more efficient than what we currently have, it's just not practical to drastically change course during spaceflight. So all spacecraft will utilise orbital mechanics to harness gravity as momentum.

The combination of these assumptions leaves us with combat based almost entirely on the skill of the combatants, but with minimal capacity to actually "fly" the spacecraft. We're a long way removed from the archetypal dogfights and flying aces. Instead, we're left with the:

Sniper Mathematician Gunslinger Hermit

The SMGH does not out-maneuver or evade their opponent, since both are all but locked into their trajectories. As such, they wouldn't really need to orient themselves. Instead, the outcome would be determined by a few things.

They would need accurate and up-to-date information on their target. If external data is susceptible to tampering, then they would need to be able to visually identify their target, and use available reference data to extrapolate the information relevant to making a shot. This would be akin to a sniper using a rangefinder based on the size of an average human, or a submariner using a ship recognition manual. I'd suggest that they have a ridiculous array of visual overlays, scopes, filters and reference manuals. Beyond that, it's up to you what other sensors can be used without risk of tampering.

The SMGH would then need to make complex ballistic calculations that account for their own velocity, the velocity of their target, and any external factors, such as gravity from nearby celestial objects. And unless their target is unaware or unable to return fire, they need to do this quickly and accurately.

You could make a distinction between ballistic weaponry and beam weaponry (if it exists in your universe) - where beam weaponry has the advantage of near-instantaneous effect, but requires line-of-sight, whereas ballistic weaponry could potentially hit a target on the other side of the solar system, as long as it didn't deviate from a predetermined course.

From those bones you build. Do your SMGH's live a life orbiting a planet for weeks/months/years at a time, honing their skills in preparedness for a single critical shot once in a blue moon? Do they hurtle after their enemy on carefully crafted parabolae and planning their shot over a period of days? Do your military heroes have a kill tally and a maximum kill distance recorded against their name?

Again, I acknowledge that this is really an alternate take on your concept.

  • $\begingroup$ The craft itself is allowed internal computers, but any connection to external sources are under threat. I like how you've turned "fighter pilot" into "sniper" too, however this question is about fighter craft. $\endgroup$
    – Aric
    Aug 3, 2017 at 10:29

The difficulty of piloting such a ship would be due to the pilot's biological inclination to be vertically oriented on a horizontal plane, and being held against it by acceleration due to gravity.

Through a neural interface, the pilot's consciousness could be disconnected from their biological sensors and integrated with the ship's own sensor hardware, designed to serve the needs of moving and being oriented freely in 3D space.

This would probably feel a lot like an out-of-body experience, where the pilot has the sensation of watching their own body or the ship it's aboard as if from a distant vantage point. This visualization would be simulated by the computer from sensor data, allowing the pilot to mentally shift their perspective arbitrarily in order to understand how things are positioned and moving in space.

It's even possible that enough training with such an interface could afford an indescribable ability to simply "understand" 3D flight, controlling the fighter as if it was a direct extension of the pilot's own body. The human brain has an incredible ability called neuroplasticity allowing it to rewire itself to process new information sources. In one real world example, the brain can be trained using feedback from a specially designed device placed on the tongue to replace data no longer available from the inner ear.

In short, my answer is that the orientation of the pilot's body doesn't need to factor into their ability to control this kind of vehicle at all.


I once played an indie space game that went for frictionless space and relied on space fighters (more like Firefly in that they were space trucks and you hauled cargo and made salvage runs. However, if you wanted to dog fight that was there too.). In either case, ships had a "correction system" that allowed for better control of the vessel. This was seen in the form of thrusters that would correct by caclulating to the speed of directional rotation and firing retro-thrusters to compensate when the pilot stopped the turn. (i.e. a ship making a right turn would fire thrusters on the right side of the ship in computer calculated bursts to counter the right turning momentum the moment the pilot stopped the turn... it was explained that the computer was quick enough and the thrusters were powerful enough to do this with little difference to atmospheric flight). The system could also be disengaged to allow for long distance travel, especially when conserving fuel and exploring areas of space that were not directly accessable by FTL. In game, most star systems had a warp gate for inter system travel and ships came with Jump Drives (essentially FTL by teleporting the entire ship to the new spot within a certain range of X, Y, and Z coordinates in space.) for shorter bursts within the system. Problem was that that the Jump Drives could only engage at maxed out speed, which ate into fuel supplies to get you up to that speed, and winning the game involved some exploration of space outside the starsystem. If one were to make a Jump Drive Jump, with the compensators off, you could essentially come out of the jump and not decellerate, effectively traveling at the highest possible speed in the game (which most engines couldn't reach outside of Jump Drive speeds), consume no fuel to maintain it, and stretch your jumps out for a much greater length of exploration before returning for a refuel. Since the game was continous space, anything within the map was accessible by flying like this given enough time and no enemy action (again, traveling too fast that they couldn't keep up with you), so I could do something for an hour, and keep an eye out for new worlds. Also was a viable way to get home if you realized you jumped too far and didn't have enough fuel to make it to safe harbors.

  • $\begingroup$ The only part of this relevant to the question is the correction system you mentioned. $\endgroup$
    – Aric
    Aug 2, 2017 at 13:48
  • $\begingroup$ Yeah... did want to show a reason as to why you might want to shut it off for fuel concerns and better describe it. I know when I'm working in my own space setting, I do use a correction system to give atmospheric like conditions and dead stops. Having a passing familiarity with real life carrier group operations, I know why they are inefficient in space compared to battleship theory. As a write, I also know they just look too damn cool, so I like to have them workable. $\endgroup$
    – hszmv
    Aug 2, 2017 at 14:54

Interesting problem! I think you have to find a balance between two goals:

  1. give the pilot as much information and as much control as possible, using as many senses and muscles as possible
  2. shield the pilot's senses from overload - especially avoid him/her getting nauseous from the rapid rotation in dogfights

For 1) you actually want to make the pilot feel any rotation of the ship. So having the pilot seat statically attached to the hull would be the way to go, to convey any rotation to the pilot via inner ear. It would be a shame to leave that sense unused.

But for 2) you'd better have the seat decoupled from the hull, so that it is moved by the computer to always face the same direction. There will still be acceleration forces, but at least no rotation. The pilot then only uses normal inputs (screen or headset, headphones, maybe a haptic sensor suit) to get information. Controlling a fighter in space with the usual two-dimensional control could be challenging, though. But at least your pilot wouldn't go crazy from the tumbling.

I maintain that today's humans are simply not biologically adapted to fighting with high-thrust ships in microgravity and three dimensions. So I think you need to define how much your humans have evolved/adapted to be good at that kind of fighting, and how much assistance/babysitting from a computer they need.

Or you could actually set it up as plot point that some pilots prefer the raw nauseating dogfighting tumble, and others prefer to touch blinking lights on a screen to command their guns.

  • 1
    $\begingroup$ Now I feel the need to re-read Brin's uplift series to see if it mentions uplifted whales/dolphins (who I could see an argument might be bioligically superior in 3d stuff) being better star fighters $\endgroup$
    – Foon
    Aug 2, 2017 at 16:40

Roller Gimbals

There's lots of reasons why you might want to re-orient the outside of the craft, but as you say spinning the pilot around is uncomfortable. So avoid it altogether.

Take the idea of a monowheel, which comprises a rotating outside while the inside remains level. Extend this into three dimensions: the outer sphere rotates, while the inside remains stationary. Every time you spin the outer sphere, the wheel between it and the inner sphere should rotate in the opposite direction.

You can then give the pilot either a very large panoramic screen or an AR headset. No actual windows. Their joystick controls will then spin the virtual display they see, and the outer sphere with all the fixed gun ports on it.


If the fighter has guns and engines capable of firing in all directions, then it has no need to rotate at all. And it's probably not possible for a human mind to has 360° spatial awareness. You would probably need multiple pilots/gunners, each minding his own sector.

If you're limited to 1 pilot and you have the ability to rotate, you should focus at least weapons in only 1 direction, just to enable pilot to focus on all of them at once. Once the "forward" direction is established, you orient the pilot facing in that direction and that's it.

FPS game controls are great analogy, especially the evolution from sophisticated System Shock layout, where you could run, look and target independently to modern shooters, where looking and targeting are fused into one control, but with more independent lateral movement. Similar thing happened with Kamov Ka-50: the helicopter has turning response good enough that the main gun no longer needs fully movable mount - it's easier to aim the whole helicopter (the gun mount still has bit of travel to do the final aiming, but not a full turret).


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .