Branson Still Doesn’t Really Understand Why SpaceShipTwo Crashed

SpaceShipTwo breaks up. (Credit: NTSB)

Virgin Galactic Founder Richard Branson was interviewed for the Jan. 30 edition of NPR’s “How I Built This” podcast. Beginning at 25:44, there’s a brief discussion of the October 2014 crash that destroyed the first SpaceShipTwo and killed co-pilot Mike Alsbury.

Branson recalls that for the first 12 hours after the accident he wasn’t sure if the SpaceShipTwo program would continue. “But, once we realized it was a pilot error and not a technical error, I was able to tell all the engineers it was nothing to do with them. And that the basic craft was sound.”

Alas, most of this explanation is wrong.

Yes, Alsbury did make a mistake by unlocking the spacecraft’s feather system early, causing the vehicle’s twin tail booms to deploy during powered ascent. Aerodynamic forces then ripped the ship apart.

SpaceShipTwo breaks up in flight as its tail booms are deployed. (Credit: Virgin Galactic/NTSB)

However, the accident was in large part about poor engineering and safety standards. It was not a sound spacecraft that Alsbury and pilot Mike Siebold climbed into that awful Halloween morning.

The NTSB investigation faulted SpaceShipOne’s builder, Scaled Composites, for both its design of the feather system and the training of the pilots.  Operating on the erroneously assumption that a pilot would never make a mistake, Scaled did not design any safeguards into the system. The vehicle that Alsbury and Peter Siebold climbed into that day had a flaw in its safety system.

Scaled’s hazard analysis relating to human error was insufficient. Rather than requiring the company to correct the shortcomings, the FAA Office of Commercial Space Transportation (FAA AST)  gave the company a waiver from the requirement. That decision was made over the objections of some of FAA AST’s own safety experts, who complained the approval process was subject to political pressure.

SpaceShipTwo’s cockpit with Mike Alsbury crashed on Cantil Road. (Credit: Douglas Messier)

With the end of the space shuttle program in 2011, FAA AST was able to recruit experienced safety experts from NASA to evaluate the first-generation space tourism vehicles. These experts had trained astronauts, understood that they made mistakes all the time, and knew how to design systems that prevented a single error from bringing down a complex spaceship. The FAA, Scaled Composites and Virgin Galactic ignored their expertise.

Virgin Galactic  took over the building of SpaceShipTwo vehicles from Scaled Composites after the accident. Branson’s engineers had to redesign the vehicle with safeguards to prevent another premature unlocking of the feather.

The changes to the feather are not trivial. So, it’s erroneous to claim SpaceShipTwo was essentially sound. The changes also bring their own risks. Engineers have to ensure the safety system they put in place to prevent a premature unlocking doesn’t malfunction when the pilots need the feature to deploy later in the flight.

Pilot error is an easy explanation. It’s also a lazy one that does not do justice to what actually happened that day. Mike Alsbury was just one part in a flawed system. He deserves better than having the entire accident laid on his shoulders.

Recommended Reading

Pilot Error is Never Root Cause, Wayne Hale

Save

Save

Save

Save

Save

Save

  • Michael Vaicaitis

    Well it was pilot error, but unfortunately for the Alsbury, as you point out, he was tasked with being part of the safety system. Lesson to be learned, whenever possible don’t put apes in charge of space vehicles. An automated flight system would also have allowed for the removal of the cockpit – making for extra passenger room or for a smaller and lighter vehicle, lowering the propulsion requirements.

  • Enrique Moreno

    I ever thought that planes are designed around “Murphy´s law”…

    I think SSII is an exception…

  • Jeff2Space

    Well said. In almost every single fatal aerospace accident, there are multiple causes. I thought that I also read that a contributing factor to this accident was pilot workload, which was far too high. This makes it more likely that mistakes will happen.

  • Jeff2Space

    Unfortunately automation is a double edged sword. With today’s technology, automation still doesn’t handle the unexpected very well. Human pilots don’t always handle the unexpected well, but they’ve got a far better chance than a simple computer algorithm does. As a person who writes computer software for a living, something unexpected happening in a computer program almost never generates a good result. In some cases, you’re quite lucky if the computer program doesn’t crash right then and there.

  • Michael Vaicaitis

    I have a software background too. Many modern fighter jets cannot fly without computers – yet there are thousands of them. Besides, what exactly are those touchscreens connected to?, flight control cables is it?. No, better to put in back-up computers and power systems than to fool the occupants into believing that they are useful as pilots.

  • JamesG

    Modern fighter jets also fall out of the sky periodically too.

  • Andrew Tubbiolo

    SSII on NPR as an example of “How I Built This”? What’s with all the press for SSII these days? The program is essentially going at a snails pace while two others are going much better but still much delayed. Who’s pushing SSII to the fore over Dragon and Starliner? And who’s responding to this push? And why? If you want an exciting program to follow, follow Falcon. It’s a fantastic pioneering project that’s really making progress on a monthly basis. Dragon and Starliner, are 2nd’s in my view, but geezh they’re like DC-3’s compared to the DH-4 “Jenny”. We do live in an exciting age, but to say SSII reflects that era is … Well ignoring the obvious and listening to con-men.

  • Jeff2Space

    While I agree with you, in part, pilots are still a vital part of controlling modern aircraft. Go read the technical details of the control system problems encountered during STS-1. Having a pilot at the controls was surely vital to landing STS-1 safely due to the flaw(s) in the flight control software.

    Pretty sure it’s in this paper:
    https://www.nasa.gov/centers/dryden/pdf/88294main_H-1912.pdf

  • Hug Doug

    It’s a different kind of vehicle that will operate in a different flight envelope than Dragon and Starliner. They’re really not in the same league.

  • Michael Vaicaitis

    A fair point, assuming computer failure is to blame for those incidents. The obvious solution is, as a backup, to put in human pilots using computer controlled touchscreens linked to computer controlled control systems, just in case the computers do fail.

  • Michael Vaicaitis

    Unfortunately for the Shuttle, 1970’s computers and software (and no doubt sensors) were obviously not up to the task. If a 2016/17 level of computing facilities had been available to them, they may have elected to go with an architecture using a propulsive vertical landing. For Musk (SpaceX) and Bezos (Blue Origin) computing is much more ingrained – perhaps the reason they chose to build space vehicles rather than aeroplanes.

  • Douglas Messier

    The podcast covered Branson’s career and the part about SpaceShipTwo covered only a couple of minutes in the 34 minute interview.

  • Douglas Messier

    In the transcripts of the NTSB interviews with FAA AST, one of the experts says the workload was the highest he had ever seen. Anywhere. I believe this safety expert had experience with the space shuttle.

    The other aspect is these pilots didn’t have a lot of experience with flying under powered conditions. Alsbury hadn’t experienced loads like that since the previous powered flight 10 months earlier. They did not have a motion simulator. The reason for the lack of flights were the engine problems they were having.

  • Jacob Samorodin

    The madmen of North Korea have 100’s of old-fashioned stick-and-rudder fighter jets. The US & SK have 100’s of fly-by-wire fighter jets within range of the 38th Parallel. Here is an ugly blood-curdling scenario: North Korea detonates a BOOSTED nuke over South Korea, The EMP pulse fries the circuitry in many of those American-built jets sitting on the runways or aprons of airbases south of the 38th. A swarm of old-fashioned MIG’s comes swooping down on those temporarily paralyzed airfields, giving North Korea temporary air superiority lasting a day or two. A million brainwashed North Korean soldiers, with 1,000’s of Soviet-era tanks and APC’s swarm out of their tunnels. What is President Trump going to do?

  • Jacob Samorodin

    I go further! Since both SS1 & SS2 architecture, ‘avionics’ and propulsion systems are derived directly and indirectly from Mr Rutan. He was never a fan of fly-by-wire systems and I believe that played a part in the stratospheric tragedy. The ground-based tragedy involving the tricky propellant can also find its cavalier handling origins with Mr Rutan as well…. Mr Rutan has blood on his hands.

  • Hug Doug

    Most all modern military equipment is hardened or shielded against EMPs. Even if most of our air power down there was fried, we still have troops, weapons, and artillery there that doesn’t require any electronics at all to do their thing. It would be pretty quick work to call on allies (say, Japan) to send some of their air power out there until we could get a carrier group there to provide off-shore air support to the troops on the ground.

  • Jacob Samorodin

    I hope you are right. Fingers crossed.

  • Hug Doug

    The USA has allies nearby that would help us out in that scenario, and military assets that could be in place within a few days, and much more that could be brought to bear within weeks. A lot of S. Korea’s infrastructure would be damaged, but militarily, N. Korea would rapidly be in an even worse position than it was at the beginning of the attack. It’s hard to see what N. Korea would have to gain from doing such a thing, anyway.

  • Andrew Tubbiolo

    That’s just what I was implying, and SSII comes out far far inferior in almost any metric of merit.

  • Andrew Tubbiolo

    I see, thanks.

  • Hug Doug

    Because they aren’t comparable. Different vehicles meant for different things.

  • Andrew Tubbiolo

    The MiG-29’s are just as vulnerable/hardened as any F-16 or F-15 which is what the USAF and SKAF fly in the region. Additionally, quite a few NORK MiG-21’s would lose systems too. Radar, IR sensors on missiles, IR sensors to warn of incoming mssiles, navigation equipment. Sure a MiG 21 would probably fly after an EMP attack, but would it believably be combat ready? I doubt their SU-25’s could hit a barn door with their targeting computers and HUD’s no longer working. If such an attack would reduce the NORKAF to MiG 15’s and 19’s then maybe we could contract out to the Confederate Air Force to hold the line until our fried electronics units that did not pass the reality test of such a scenario are replaced with fresh units.

  • Andrew Tubbiolo

    Also consider that your EMP attack would induce amazing currents into the Chinese national power grid. The Soviets lost an amazing fraction of the electrics in their machine tools all along major grid lines that passed through Siberia during their version of Starfish Prime. Think of the basic physics an EMP, the closed path integral of E along dl, where dl is swept through distances as long as continents, and E is being pumped by a nuclear bomb. Literally, astronomical currents would be induced. You’ll fry every CNC machine tool on the East Coast of China and probably into the interior as well. What Would Mao Do?

  • JamesG

    Lots of contingency plans for that. Immediately, mostly involving Carrier Battle Groups, B1s, B2s, and B52s, the JSDF, and then …. probably most of the rest of the world signing on the beat the stuffing out of N. Korea.

    Which is why hasn’t and never will happen. The Kims want nukes and delivery systems not as an offensive weapon as we are told by our own Mil-Ind-Complex looking to sell new toys, but as a “dead-man’s switch” to make digging them out of Pyongyang more expensive than its worth.

  • Jacob Samorodin

    It is not my EMP attack! Do you seriously think the North Korean leaders think as rationally as you and I do? Common! If you gave the North Korean leader the info (you posted to me) to him face to face, you would end up in front of a firing squad.

  • Jacob Samorodin

    The North Korean leaders don’t think like you or I, but we can draw some conclusions about how they think. They think they could some day soon carry out a surprise attack against the south. That’s debatable. They know their airforce won’t last long in any confrontation, knowing Japanese-based fighters will come to the south’s rescue within a day or two of any preemptive air-to-ground strike against the south’s airfields. I guess that’s the only thing MIG’s could do successfully is TRY to hit the aircraft on the ground…. And of course, President Trump could end things…quickly….by having Pyongyang vanish into radioactive dust within a minute of a Trident missile reaching ground-zero point above the city before it disappears. I hope he would find a better response.

  • Andrew Tubbiolo

    I would say the NORK leadership thinks like most corporate boards with a heavy salting of something along the lines of the Yakuza thrown in the mix. When it comes to staying alive and in power, they are mostly rational. The balance of forces of the South vs the North are pretty stark. The South has over 1500 M1A1 class Main battle tanks vs 3000 or so T-55 class machines. The South has owned approx 100 T-80’s and even a few T-95’s to practice against in the event the NORKs get their hands on modern armor. If you understand the differences between a M1A1 and a T-55, I need not say more about what would happen at the 48th parallel if the North decides to go South. The gap is huge in every metric, including population from which to draw combatants. The NORKs know this, they’re not going South. The secret to NORK survival has been the fact that the South is having such a great time being themselves, they’re not really interested in lifting their Northern cousins out of the 1880’s. And China needs them there to act as a brake on the S Korean economy. If the NORKs were as insane as you think they are, they would have tried to move South in the 90’s or early 2000’s before Papa Kim died and the S Korean economy was just starting rocket away from the North’s. Or after Jong Il, had consolidated power and his health began to decline. More so than his grandfather, and father, Jong Un answers to the divisional commanders of those tank, infantry, and artillery divisions. He has to keep them in power, well fed, equipped, and rich. They know their life’s work/inheritance would be smashed by the forces of the South if they actually tried to use those forces in combat against the South. They simply won’t let Jong Un do something like that if he were truly inclined to do so.

  • patb2009

    They do what they did in the Korean War. Depend upon Carrier aviation,
    to provide backup cover and they use Fighter jets transited up from guam and Okinowa out of Japan. They use Air Defense artillery and missiles and count on the low skills of the KPAF to reduce effectiveness.

  • patb2009

    What Would Mao Do? Decompose.

    The better question is what would Xi Jinping do?

  • patb2009

    The only thing the PRK has over the RoK is sheer manpower.
    There are a lot of them. But they are mostly illfed draftees.

  • publiusr

    I can’t help bout wonder if this is a case of trying to do too much with too small a craft.

    I think Branson and Allen need to combine efforts.

    Stratolauncher could release a larger, more conventional X-34 type sub-orbital rocket-plane with a few more paying passengers. This is what a Tier-3 was supposed to look like anyway. Two pilot stations–and yes–a flight engineer.

  • James

    Actually there have been quite a number of incidents where heavy computer controlled aircraft have messed up massively and had to be taken over by pilots. Its one of the problems that has popped up with some overseas airline corps who use crappy pilots because they have really nice computer controlled systems.

    In reality the captain in a airliner is just that. He is a captain of a ship. The only difference is the Airliner itself is flown mostly by a computer pilot.