SpaceShipTwo: connect flight phases and human factors

SpaceShipTwo developed the air-launched suborbital approach for a larger vehicle. WhiteKnightTwo carries the rocket plane before release; powered ascent, a coast, atmospheric return and a runway landing follow as distinct phases. NASA documented the first SpaceShipTwo powered test in April 2013. A powered test flight is an engineering milestone, not automatic proof that every later part of the intended flight envelope has been demonstrated.

The vehicle’s feather system changes its configuration for atmospheric return, increasing drag and helping control its attitude. Configuration suitable for one phase can be unsuitable for another. This makes the timing and conditions of transitions part of the design problem. Engineers need to understand both normal commands and what happens if a command occurs at an unintended time.

During a powered test on October 31, 2014, the first SpaceShipTwo broke apart. The copilot died and the pilot was seriously injured. The NTSB found that the feather was unlocked too early and then extended without a deployment command under the flight loads. Unlocking a mechanism and commanding it to move are different actions, even when external forces can connect their consequences.

The investigation’s probable-cause finding centered on Scaled Composites’ failure to anticipate and protect against a single human error creating a catastrophic hazard. Reducing that conclusion to a story about an individual pressing the wrong control misses the engineering lesson. A system’s safeguards, workload assumptions and training conditions must be evaluated together. The finding concerns the documented accident and should not be replaced by speculation about the crew’s intentions.

For a hypothetical vehicle, imagine a latch that must remain closed during high loads. A written procedure tells an operator when to release it, but the design review must also ask what happens if it is released early. An interlock, clearer state information or changed task sequence might reduce the hazard; each proposed measure needs verification against its own possible failures.

Human factors — open definition work connects hardware, displays, procedures and the environment in which people act. Vibration, workload and unfamiliar conditions can change performance. Testing only an ideal sequence leaves important behavior unexplored. Later vehicles and flights need their own evidence of the changes made and conditions tested. The purpose of studying an accident is to improve the entire system while preserving an accurate, respectful account of what happened.