Why is skylon unmanned?

[Nuke]

lots of plane crashes have resulted from pilots improperly configuring the autopilot. one instance where a pilot set the decent rate in meters thinking it was in feet, and the plane subsequently flying itself into the ground. also you got pilots relying on their instrumentation even though they could look out the window and tell that something wasnt right. 'the computer says were fine so were...' *plane crashes*

automated flight is better off being implemented in an all or nothing sort of way. though i think i saw something on the news a few months ago about drone crash rates being like four times greater than piloted crash rates. i dont think that applies to rockets at all, but spaceplanes is a different matter. i dont think you are doing any kind of exotic maneuvering with a spaceplane that you would normally do in a military drone that would endanger the craft. so its an apples to oranges comparison anyway.

Edited January 12, 2014 by Nuke

[softweir]

Not quite. They cannot take off on their own and the auto flight systems do not have control of the flaps or landing gear on any of the Boeing or Airbus aircraft that I am familiar with. Autopilots also cannot interact with ATC. An autopilot is just a tool to reduce workload. It allows the crew to focus on what humans are good at but the pilots still need to manage the autopilots through the various phases of a flight.

Then again, the Soviet Buran vehicle managed a 100% automated flight from launch, through orbit to automated landing. (Assuming the technicians didn't cheat and use remote control at any stage.) It can be done - if sufficiently talented and resourced coding teams get to do it!

I wonder if those guys might be available to program Skylon?

[Nibb31]

Not only it can be done, but automatic landing is mainstream these days. No need to recruit a bunch of retired Russian programmers.

All sorts of drones take off and land every day, including on aircraft carriers and in bad weather. The X-37B reenters and lands automatically, and so will DreamChaser. The reason airliners still have pilots is for safety, insurance, and PR reasons.

Edited January 12, 2014 by Nibb31

[Seret]

also you got pilots relying on their instrumentation even though they could look out the window and tell that something wasnt right. 'the computer says were fine so were...' *plane crashes*

That's what they're trained to do. Your eyes and ears will trick you, your instruments won't. Trainee pilots are actually taken up and instrument flying hoods or visors are fitted to block out their view to force them to do it on instruments.

[rpayne88]

That's what they're trained to do. Your eyes and ears will trick you, your instruments won't. Trainee pilots are actually taken up and instrument flying hoods or visors are fitted to block out their view to force them to do it on instruments.

This is a bit off topic, but you're not exactly right. In instrument meteorological conditions (IMC, AKA bad weather,) pilots are told to rely on their instruments. They also need an instrument rating to do that. Otherwise, (with the exception of flying in class A airspace) flying consists of avoiding IMC and flying under visual flight rules (VFR) using the horizon as a reference. The only time a VFR pilot relies solely on his/her instruments is in the case of inadvertent flight into IMC, in which the procedure is to maintain altitude, make a 180 degree turn, and fly straight and level until they can see the horizon again. That is the only reason student pilots occasional fly "under the hood."

Edited January 12, 2014 by rpayne88

[PakledHostage]

rpayne88 already gave a good and knowledgeable correction to Seret, so I'm going to tackle the point about UAV safety. As Nibb31 pointed out, airliners are not as automated as UAVs for safety reasons, not because it can't be done. The accident rate for UAVs is very high, even though most of the UAVs that achieve the dubious honour of being included in the statistics are operated and maintained by professional military crews. An NBC News story cites a reference that claims that UAV's have an accident rate 30 to 300 times higher than general aviation aircraft (i.e. private pilots flying mostly piston powered aircraft). Autoland systems on airliners are "mainstream", but they are only safe because they are monitored closely in real time by human crews. The ground based and airborne segments of the system also continually need to be tested to remain qualified.

But even despite the demonstrated risks, it makes sense to design Skylon to fly itself. It is much more "cutting edge" than an airliner. If they succeed at all, the margins are going to be very slim. It is likely that attempting to man rate the system, at least initially, could make the difference between success and failure of the project.

[Seret]

This is a bit off topic, but you're not exactly right. In instrument meteorological conditions (IMC, AKA bad weather,) pilots are told to rely on their instruments. They also need an instrument rating to do that. Otherwise, (with the exception of flying in class A airspace) flying consists of avoiding IMC and flying under visual flight rules (VFR) using the horizon as a reference. The only time a VFR pilot relies solely on his/her instruments is in the case of inadvertent flight into IMC, in which the procedure is to maintain altitude, make a 180 degree turn, and fly straight and level until they can see the horizon again. That is the only reason student pilots occasional fly "under the hood."

Sure, and of course there are plenty of folks who operate predominantly or totally in VFR. I just wanted to point out that relying on instruments wasn't at all alien to pilots, and not something to be pointed out as a weakness.

[NovaSilisko]

Disregarding all previous posts to answer the OP's question directly: Humans are big, heavy, fragile things, and most of the things launched with skylon are going to be satellites. As others have stated, maybe for crewed flights you'd want... well, crew on board, to mitigate any mishap but that can be integrated into the capabilities of the passenger cargo pod if it even is necessary at all. Pretty much everything on skylon that's not the cargo bay, an engine, or a wing, contains propellant. If you wanted to redesign it so it had windows at the front for crew in the cockpit or something, you'd have to totally redesign the interior of the vehicle (and in the process you'd probably screw over the center of mass). Even then, crewed flight is entirely theoretical for skylon... as is skylon itself, but I digress.

Now, no longer disregarding previous posts (except some I probably missed): I'm not sure about the whole "we need humans to monitor it" thing. Unmanned aircraft have come a long, long way since the shuttle was first conceived... the Global Hawk UAV, for instance, is capable of wholly autonomous flight. As in, the operator (you really can't call him a pilot) just has to press the take off button, and then the aircraft follows a series of GPS waypoints to the target, then comes home.

[Nuke]

Sure, and of course there are plenty of folks who operate predominantly or totally in VFR. I just wanted to point out that relying on instruments wasn't at all alien to pilots, and not something to be pointed out as a weakness.

i was just trying to point out the issue of 'automation addiction', cases where dependance on automation caused a crash. a well trained, experienced pilot should have no problem operating in either vfr or ifr. in the context of space planes heavy automation is the only way to fly, manned controls are at most an emergency, last chance or your dead, omg were all gonna die, backup.

[Seret]

i was just trying to point out the issue of 'automation addiction', cases where dependance on automation caused a crash.

I don't disagree that it could happen, but on balance automation prevents a lot more crashes than it causes.

in the context of space planes heavy automation is the only way to fly, manned controls are at most an emergency, last chance or your dead, omg were all gonna die, backup.

As with all unmanned spacecraft, you're still going to have a human in the loop, even if their only meaningful intervention is hit the big red range safety button. Being a spaceplane though I would imagine Skylon (if it ever flies) would have a few more abort options throughout much of it's flight than most spacecraft, so the meatsack doing the machine minding would have a meaningful role in an emergency.

[magnemoe]

i was just trying to point out the issue of 'automation addiction', cases where dependance on automation caused a crash. a well trained, experienced pilot should have no problem operating in either vfr or ifr. in the context of space planes heavy automation is the only way to fly, manned controls are at most an emergency, last chance or your dead, omg were all gonna die, backup.

Unmanned is not the same as totally automated. It would be automated with remote control, downside of this over crewed is the lack of situation awareness, the lack of situation awareness is that kills UAV.

However Skylon will operate pretty different from an predator UAV who is an small low performance aircraft who fly a lot between the mountains in Afghanistan.

With Skylon mission profile situation awareness will be of limited use. take off, burn for orbit, in orbit it behaves like any other remote controlled craft.

You set the deobit burn and adjust, during deorbit you loose communications so it has to run totally automated, until into the lower atmosphere.

It would probably be set up to could land totally automated but you can also do this by remote.

[Seret]

From what I've read of Skylon its ground pressure would be pretty high, so it would only be able to operate from specially reinforced runways. That means you'd be operating inside a known and predictable flight corridor, likely out of a single location (although I suspect they'd like to have some divert options if possible).

[rpayne88]

Up to now, we've all been talking about using a human pilot in the event of an emergency to land the thing. What about other stuff such as life support failures and what not? If it turns into the equivalent of the airlines with untrained individuals going up, how are you going to talk them through improvising a solution to, say, an air scrubber? Its hard to get machines to improvise, never mind if some power failure means you have to save what power you have left for EDL.

[DerekL1963]

I don't disagree that it could happen, but on balance automation prevents a lot more crashes than it causes.

Citation needed.

[Themohawkninja]

Citation needed.

It's no .edu study, but it seems legit: http://planecrashinfo.com/cause.htm

[Thiel]

Up to now, we've all been talking about using a human pilot in the event of an emergency to land the thing. What about other stuff such as life support failures and what not? If it turns into the equivalent of the airlines with untrained individuals going up, how are you going to talk them through improvising a solution to, say, an air scrubber? Its hard to get machines to improvise, never mind if some power failure means you have to save what power you have left for EDL.

If it ever turns into a commercial people carrier, which I highly doubt it will, then you build in redundant system. Just like modern airliners do.

[Seret]

Citation needed.

Sure:

From CAST:

Automation has contributed substantially to the sustained improvement in air carrier safety around the

world.

From Airbus:

Correct use of automated systems reduces workload and significantly improves

the flight crew time and resources for responding to:

-An unanticipated change (e.g., ATC instruction, weather conditions, …); or,

-An abnormal or emergency condition.

That claim doesn't really need a documentary source though. Just look at the avionics package fitted to modern aircraft. More and more of the crew's workload has been automated over the years, in the name of efficiency and safety. The flight crew of an airliner used to be four, now it's two. We've automated the navigator and flight engineer out of existence, and much of the pilots' workload is automated now too.

They wouldn't be doing this if the systems performed worse than a human.

[magnemoe]

From what I've read of Skylon its ground pressure would be pretty high, so it would only be able to operate from specially reinforced runways. That means you'd be operating inside a known and predictable flight corridor, likely out of a single location (although I suspect they'd like to have some divert options if possible).

Takeoff pressure will be high, might be an idea with an trolley for launch so you can use lighter landing gear.

Note that long distance flights can not land with their takeoff weight, they need to dump fuel before landing.

Today they burn the fuel instead of dumping it if its not an emergency like if landing gear can not be retracted you will end up circling and burn fuel until you can land.

Wonder if it would be an idea to fly the skylon to an runway close to equator and refill it before burning for orbit, then fly it back to base?

[PakledHostage]

Takeoff pressure will be high, might be an idea with an trolley for launch so you can use lighter landing gear.

I think that is a really interesting idea.

Note that long distance flights can not land with their takeoff weight, they need to dump fuel before landing.

Pilots have the option of landing overweight, but the aircraft must be inspected for structural damage after landing. It isn't that onerous an inspection but it must be done before further flight, and a more detailed inspection will be required if the initial inspection finds any warning signs.

Also, fuel dumping systems aren't always installed. When they are, dumping usually takes place in airspace that has been designated for the task. Swissair flight 111 had been heading towards such an area when they crashed. They were under the mistaken belief that their problem wasn't urgent and were going to dump fuel in the designated area rather than land overweight or seek permission to dump fuel anywhere. As the situation became more dire, they were given permission to dump fuel before reaching the designated area, but by then it was a moot point.

[magnemoe]

U2 have use the trolley launch sometimes, if you want to go real advanced you can put engines on it to help with takeoff.

Understand overweight landing, also know that they prefer to burn fuel today of environmental reasons, previously it was dumped.

I was in an flight from the US to Europe then they got an problem with the navigation system, think they lost one system and an backup was required for transatlantic flight, engines was very load and we was told this wast to burn fuel before landing. 10-20 years ago it would be dumped.

Was in another flight who was just 2000 km where the nose wheel did not retract and it could just turn around and land as the plane was light enough.

Even smaller planes for shorter flights lack the fuel dump abilities as they sometimes has to circle to burn fuel if they has to land without wheels who is an obvious emergency.

Edited January 14, 2014 by magnemoe

[Thiel]

U2 have use the trolley launch sometimes, if you want to go real advanced you can put engines on it to help with takeoff.

So did many early rocket fighters and the V1 flying bomb. The former tended to favour skids as the simplest form of landing gears though it did suffer some in the comfort department. The latter obviously didn't need any landing gear. The V1's trolley was further propelled by liquid fuelled rocket. (HTP and Potassium permanganate)

[Wahgineer]

Skylon was meant to be unmanned since the beginning. If a skylon experiences some sort of failure, all that needs to happen is for the passenger unit/palette to burst out of the cargo bay and parachute to safety. All the control systems, life support, and space necessary to allow the skylon to be mannned would negate the moderate load/ high turnaround/ low cost advantages it will have over most other LEO launch systems.

[Mmmmyum]

I'd be more than happy without a pilot, it gets rid of human error. Also like others have said people tend to add a lot of weight to a spacecraft and they aren't really needed.

[DerekL1963]

I'd be more than happy without a pilot, it gets rid of human error.

Except, of course, for all the humans who designed, built, and serviced the vehicle.

(And a note to Seret, while your soundbites sound impressive, they fail to support your contention that automation has prevented more crashes than it has caused. Especially in light of the dramatic rate of UAV crashes.)

[Seret]

And a note to Seret, while your soundbites sound impressive, they fail to support your contention that automation has prevented more crashes than it has caused. Especially in light of the dramatic rate of UAV crashes.

I was talking about automation in manned aircraft.