Somewhere in the shuttle design is a lesson about not cramming too much into the MVP. Many years ago, in some high school engineering camp, I remember a university professor who had been involved in the STS design talking about feature creep in the shuttle program in very uncomplimentary terms.
According to the prof, the shuttle as originally envisioned would have had about 1/2 the cargo volume, and less throw weight, but in return a higher orbit at engine cut-off, much better serviceability, and a simpler design. In order to get the support of the military, the cargo requirement was pushed to what was just-barely-heroically-possible engineering in order to launch the military satellites desired by the spooks.
This decision compromised the program in every way imaginable. The schedule stretched out. The launch-pad weight ballooned, forcing the solid booster design into barely-possible territory. The original goal of "recycle for weekly launches" became a pipe dream. The re-entry heat shield design became very complex. The unit cost ballooned. The cost per launch ballooned.
The right answer all along would have been to tell the spooks: "No, damn it, design satellites that fit the payload bay or take a hike." The spooks would have had many years to refine and compact ever-better satellites. We paid a high price for poor program management decisions at the top.
The lesson I take out of it: As engineers, we need to think about risk budget at all times. Pushing the envelope on something is probably necessary to beat the competition. Pushing the envelope on everything is poor risk management. As an engineering manager, be very intentional about where risk is allowed. Concentrate your management efforts on de-risking the risky, and delegate the low-risk to clear your mind for managing the unpredictable.
I will admit that I have not always done that, and have collected the scars that I deserved.
In the end, feature de-creep turned out to be a bigger safety problem than feature creep. The early STS design proposals called for a titanium airframe which could have withstood reentry temperatures. But to reduce costs and speed up manufacturing they switched to an aluminum airframe covered with ceramic tiles. Those tiles were fragile, leading to one crash and several near misses (like this article).
Also the performance of spy satellites is dictated by optics. In order to get sharper images you need larger lenses. That's just physics. Smaller will always be worse.
Regarding the feature creep Pentagon Wars come to mind :)
Another answer is modularity of design, like Falcon Heavy. Though granted it is much harder for a spaceplane concept and may as well be an ultimate feature creep.
Launching a (classifies) satellite into a useful-for-spying orbit is one level of secret, although clearly something which could be done with unmanned rockets (as was done later).
One of the super secret missions they used to sell the shuttle program was capture (and theft, or tampering and replacement?) of foreign in-orbit satellites. There isn’t any solid evidence this was ever done, but it was a unique capability of something like the shuttle.
>> One of the super secret missions they used to sell the shuttle program was capture
No. That is the easier-to-swallow purpose. The logistics of capturing an enemy sat, one that might defend itself by changing orbit, are ridiculous.
What shuttle could do was bring friendly sats back. Why? Well it would have to be something that you couldn't let burn up in the atmosphere but neverthless needed to come home. Shuttle could have serviced space-based nuclear weaponry, the one type of friendly sat that couldn't be destroyed in the atmosphere or sent into a disposal orbit.
A fleet of nuclear-tipped reentry vehicles in orbit would have required constant up-and-down servicing. Russia saw the US biulding a vehicle capable of servicing such a fleet. Fearing a capability gap, Russia developed Buran. But when various treaties and understandings mooted space-nuke concepts, Buran lost its purpose and was parked. Shuttle continued.
Other aspects of shuttle also came from never-flew military missions. Those giant wings were to permit a polar-orbit overflight of russia with return to the launch center, a cross-range capability. It didn't need such massive wings.
> Launching a (classifies) satellite into a useful-for-spying orbit is one level of secret, although clearly something which could be done with unmanned rockets (as was done later).
(As was done before, during and after the space shuttle program.)
In this case the story says the satellite required ad hoc human intervention in order to become operational which does support the idea that the STS was needed.
However given the amount of propaganda around both the STS program itself and of course the NRO and other intelligence agencies it's difficult to know what to believe, and I pretty much agree with the opinion (yours too I believe) that the STS was a significant drag on the US space program.
(What's the chance that purpose built "satellite retrievers" might already have been constructed? The current crop of micro satellites would be seemingly easier to retrieve and return to earth using a "corona-style" mechanism (drop and collect in air via aircraft) -- something impossible with the older refrigerator- or bus-sized ones.
> One of the super secret missions they used to sell the shuttle program was capture
That would work once. After you successfully capture one satellite your adversary would add a special explosive payload to a subsequent satellite and then "leak" a salacious detail about said satellite to make you attempt to capture it. One shuttle braking up in orbit later and you'd never use it for that purpose again.
> Flight controllers were convinced from the grainy images that the damage was not severe and that the crew were mistakenly seeing damage in conditions of poor lighting.
I can't help but wonder if they had to say that because there was no real way to either fix the shuttle or retrieve the astronauts.
> If the heat shield was damaged, it could spell disaster during the fiery return to Earth and Gibson’s was instructed to use RMS cameras to acquire imagery.
Why wasn't this SOP until after we lost another shuttle crew?
Supposedly (I can't find a reliable source), Neil Armstrong was once asked what he would have done with his final hours on the moon if the ascent engine had failed and they had been stranded. His response was that he'd have spent his final hours trying to fix the engine.
The point is that there's always something to try, even if it's improbable. Pilots are told to fly as far into the crash as possible, because it never helps to give up, and there might still be something you can do.
It's possible that they chose to say that because they thought there was no way to save the astronauts, but they definitely did not have to. And if that's really how it was, they definitely shouldn't have.
Supposedly NASA was aware of the tile damage to Columbia before re-entry in 2003, but purposefully didn't tell the astronauts because they believed there was indeed no way to fix it; the Shuttle's TPS tiles are right on the bleeding edge of being capable of protecting the craft, just barely thick enough to do their job (thin to save weight) and completely unable to resist impacts. Supposedly the philosophy was such that NASA management would rather the astronauts be completely oblivious to the impending failure during re-entry, carry out a successful mission, and thus experience only a short period of pure horror during re-entry, as opposed to informing them there was nothing to be done and that re-entry would kill them.
I have never understood NASA management. It's been responsible for both Shuttle disasters (pushing Challenger to launch despite unsuitable weather). NASA demonstrated what I think is humanity's finest engineering-your-way-out-of-a-problem with Apollo 13, which couldn't benefit from satellite imaging, robot arms or other inventions since, but they seem pretty complacent after that.
> Pity they didn’t learn the lesson and maybe Columbia in 2003 could have been saved. But again, mission controllers and engineers ignored the information as “insignificant".
That was in the comments. That is the best part of the article. We have a long road to go about learning from things that worked ok by pure luck. But, as usual, we barely learn from mistakes...
I don't think any engineers deemed any available information about Columbia as "insignificant", not that there was very much information to go on due to lax execution by camera tracker teams. Management on the other hand was... not effective.
I think that's really the crux of your statement. Humans seem to be terrible at working together in large organizations, or at least running them. Organizations seem to learn from mistakes only temporarily, or at all (as you pointed out). Consider the self-driving car group at Uber as a recent example.
"But, as usual, we barely learn from mistakes... "
As in, for example, public warnings by countless actual hackers back in the 80s and 90s about how vulnerable most US systems were. (For which, in return, they were persecuted and prosecuted. With no apologies as said warnings came to deserved fruition.)
"One tile on the shuttle's belly near the nose was completely missing and the underlying metal - a thick mounting plate that helped anchor an antenna - was partially melted. In a slightly different location, the missing tile could have resulted in a catastrophic burn through."
> During their training, they earned the nickname “Swine Flight” from the astronaut office secretaries, and were even given novelty pigs’ snouts, as a result of Gibson’s penchant for making animal-like snorts whenever attractive women were in the vicinity.
Gross. I'm glad culture, as well as technology, has evolved since the 80's.
It has, but let's not confuse that with it being good now. A lot of this sort of thing still goes on at a lot of very high profile places. I've personally seen it happen--and worse!
I am wondering why the images of the damage had to be encrypted - what sort of secret about the mission could they reveal? My only guess, which I don't find very convincing, is that it would reveal some modification that had been made to the robot arm, or its camera, for the purpose of the mission.
Note the benefit of having a shuttle - the ability to fix the satellite after launching. Too bad the shuttle design (attached to the tank etc) is not ideal.
It was almost certainly a matter of policy (pre-STS-51L, there was still quite a strong push towards the shuttle taking over all US launches) rather than strict necessity. Some flights (STS-41D, for instance) launched civilian comsats which surely can't have had any terribly exotic requirements.
[+] [-] dbcurtis|7 years ago|reply
According to the prof, the shuttle as originally envisioned would have had about 1/2 the cargo volume, and less throw weight, but in return a higher orbit at engine cut-off, much better serviceability, and a simpler design. In order to get the support of the military, the cargo requirement was pushed to what was just-barely-heroically-possible engineering in order to launch the military satellites desired by the spooks.
This decision compromised the program in every way imaginable. The schedule stretched out. The launch-pad weight ballooned, forcing the solid booster design into barely-possible territory. The original goal of "recycle for weekly launches" became a pipe dream. The re-entry heat shield design became very complex. The unit cost ballooned. The cost per launch ballooned.
The right answer all along would have been to tell the spooks: "No, damn it, design satellites that fit the payload bay or take a hike." The spooks would have had many years to refine and compact ever-better satellites. We paid a high price for poor program management decisions at the top.
The lesson I take out of it: As engineers, we need to think about risk budget at all times. Pushing the envelope on something is probably necessary to beat the competition. Pushing the envelope on everything is poor risk management. As an engineering manager, be very intentional about where risk is allowed. Concentrate your management efforts on de-risking the risky, and delegate the low-risk to clear your mind for managing the unpredictable.
I will admit that I have not always done that, and have collected the scars that I deserved.
[+] [-] sgt101|7 years ago|reply
[+] [-] nradov|7 years ago|reply
https://history.nasa.gov/SP-4221/ch8.htm
Also the performance of spy satellites is dictated by optics. In order to get sharper images you need larger lenses. That's just physics. Smaller will always be worse.
[+] [-] trhway|7 years ago|reply
Another answer is modularity of design, like Falcon Heavy. Though granted it is much harder for a spaceplane concept and may as well be an ultimate feature creep.
[+] [-] rdl|7 years ago|reply
One of the super secret missions they used to sell the shuttle program was capture (and theft, or tampering and replacement?) of foreign in-orbit satellites. There isn’t any solid evidence this was ever done, but it was a unique capability of something like the shuttle.
[+] [-] sandworm101|7 years ago|reply
No. That is the easier-to-swallow purpose. The logistics of capturing an enemy sat, one that might defend itself by changing orbit, are ridiculous.
What shuttle could do was bring friendly sats back. Why? Well it would have to be something that you couldn't let burn up in the atmosphere but neverthless needed to come home. Shuttle could have serviced space-based nuclear weaponry, the one type of friendly sat that couldn't be destroyed in the atmosphere or sent into a disposal orbit.
A fleet of nuclear-tipped reentry vehicles in orbit would have required constant up-and-down servicing. Russia saw the US biulding a vehicle capable of servicing such a fleet. Fearing a capability gap, Russia developed Buran. But when various treaties and understandings mooted space-nuke concepts, Buran lost its purpose and was parked. Shuttle continued.
Other aspects of shuttle also came from never-flew military missions. Those giant wings were to permit a polar-orbit overflight of russia with return to the launch center, a cross-range capability. It didn't need such massive wings.
[+] [-] gumby|7 years ago|reply
(As was done before, during and after the space shuttle program.)
In this case the story says the satellite required ad hoc human intervention in order to become operational which does support the idea that the STS was needed.
However given the amount of propaganda around both the STS program itself and of course the NRO and other intelligence agencies it's difficult to know what to believe, and I pretty much agree with the opinion (yours too I believe) that the STS was a significant drag on the US space program.
(What's the chance that purpose built "satellite retrievers" might already have been constructed? The current crop of micro satellites would be seemingly easier to retrieve and return to earth using a "corona-style" mechanism (drop and collect in air via aircraft) -- something impossible with the older refrigerator- or bus-sized ones.
[+] [-] cptskippy|7 years ago|reply
That would work once. After you successfully capture one satellite your adversary would add a special explosive payload to a subsequent satellite and then "leak" a salacious detail about said satellite to make you attempt to capture it. One shuttle braking up in orbit later and you'd never use it for that purpose again.
[+] [-] dasmoth|7 years ago|reply
On-orbit servicing of friendly recon sats was considered pretty seriously, though: http://www.thespacereview.com/article/3390/1
[+] [-] JustSomeNobody|7 years ago|reply
> Flight controllers were convinced from the grainy images that the damage was not severe and that the crew were mistakenly seeing damage in conditions of poor lighting.
I can't help but wonder if they had to say that because there was no real way to either fix the shuttle or retrieve the astronauts.
> If the heat shield was damaged, it could spell disaster during the fiery return to Earth and Gibson’s was instructed to use RMS cameras to acquire imagery.
Why wasn't this SOP until after we lost another shuttle crew?
[+] [-] mikeash|7 years ago|reply
The point is that there's always something to try, even if it's improbable. Pilots are told to fly as far into the crash as possible, because it never helps to give up, and there might still be something you can do.
It's possible that they chose to say that because they thought there was no way to save the astronauts, but they definitely did not have to. And if that's really how it was, they definitely shouldn't have.
[+] [-] gargravarr|7 years ago|reply
I have never understood NASA management. It's been responsible for both Shuttle disasters (pushing Challenger to launch despite unsuitable weather). NASA demonstrated what I think is humanity's finest engineering-your-way-out-of-a-problem with Apollo 13, which couldn't benefit from satellite imaging, robot arms or other inventions since, but they seem pretty complacent after that.
Source: https://en.wikipedia.org/wiki/Space_Shuttle_Columbia_disaste...
[+] [-] woliveirajr|7 years ago|reply
That was in the comments. That is the best part of the article. We have a long road to go about learning from things that worked ok by pure luck. But, as usual, we barely learn from mistakes...
[+] [-] ubertakter|7 years ago|reply
I think that's really the crux of your statement. Humans seem to be terrible at working together in large organizations, or at least running them. Organizations seem to learn from mistakes only temporarily, or at all (as you pointed out). Consider the self-driving car group at Uber as a recent example.
[+] [-] 8bitsrule|7 years ago|reply
As in, for example, public warnings by countless actual hackers back in the 80s and 90s about how vulnerable most US systems were. (For which, in return, they were persecuted and prosecuted. With no apologies as said warnings came to deserved fruition.)
[+] [-] JustSomeNobody|7 years ago|reply
[0] Yes, safe is relative. Are they safer than humans is one measure. But should we stop there? No. Can we truly even get there? ...
[+] [-] mannykannot|7 years ago|reply
https://spaceflightnow.com/shuttle/sts119/090327sts27/
[+] [-] mr_overalls|7 years ago|reply
Gross. I'm glad culture, as well as technology, has evolved since the 80's.
[+] [-] moron4hire|7 years ago|reply
[+] [-] mannykannot|7 years ago|reply
[+] [-] Diederich|7 years ago|reply
[+] [-] hindsightbias|7 years ago|reply
[+] [-] vermontdevil|7 years ago|reply
[+] [-] java-man|7 years ago|reply
[+] [-] radicaldreamer|7 years ago|reply
[+] [-] dasmoth|7 years ago|reply
[+] [-] gsnedders|7 years ago|reply