April 17, 2023. SpaceX encountered a pressurization issue with the massive Super Heavy booster. Engineers couldn’t quite figure it out in time. With less than 10 minutes on the countdown clock, the Starship launch was aborted.
Three days later, Starship blasted off on schedule at 8:33am Central Time. There was a lengthy pause during the countdown at T-minus 40 seconds. Spectators waited in anticipation. Before long, the countdown continued uninterrupted until liftoff.
Not all the 33 Raptor engines fired up at the start. Three engines failed to ignite.
After 40 seconds, one engine shut down, leaving 29 raptor engines pushing Super Heavy upwards. Two more engines went off later.
Nevertheless, Starship achieved a maximum speed of 2157 km/h with 27 engines firing.
At T+ 2m 50s, Starship started flipping to prepare for stage separation. That’s went things started to go wrong, or at least, not according to plan. Instead of separating, Starship continued to rotate for 3.5 times before exploding.
Was this a failure? It depends on whose perspective you follow. Elon Musk tweeted, “Congrats SpaceX team on an exciting test launch of Starship! Learned a lot for next test launch in a few months.”
For Elon, this was a test launch to gather data and lessons learned. A smart failure.
The official SpaceX Twitter account surmised, ""With a test like this, success comes from what we learn, and today’s test will help us improve Starship’s reliability as SpaceX seeks to make life multi-planetary.”
NASA Administrator Bill Nelson told CNN that the launch of SpaceX's Starship rocket is "a good first step… they got through the first stage of this big monster rocket. That’s a real accomplishment. We’ll get a report on what happened to the second stage, but I’m very encouraged that they’ve gotten along this far.”
Before Thursday, Musk set expectations by saying, “success is not what should be expected…That would be insane.”
Based on that remark, the April 20 explosion was a failure. But what kind of failure?
Let’s evaluate it from a project management perspective.
The goal of the test flight was to get to near orbital speeds and make a partial lap of the planet.
Orbital speed is 7.9km per second or 224,676 km/h. The Super Heavy achieved a maximum speed of 2,157 km/h or 0.96% or orbital speed. By that measure, it was a failure. There was no way to achieve anything close to orbital speed without the second stage firing.
What was the minimum threshold of success? Yes, clearing the launchpad, collecting lots of data, and deriving lessons learned from the data.
Compare the Starship explosion against NASA’s successful launch of the Artemis SLS rocket on November 16, 2022. Two solid rocket boosters (SRBs) performed exactly as planned. The core stage with 4 RS-25 engines fired as planned, as did the upper stage single RL-10 engine which powered Orion all the way to the moon and back. The Orion space capsule successfully re-entered Earth’s atmosphere and landed intact in the Pacific Ocean on December 11, 2022. Now, that’s a successful project that met its goal.
From that standpoint, the best that can be said of Starship’s test launch was it’s successful failure! How’s that for an oxymoron? Is there even such a phrase the project manager’s lexicon?
Is this an agile, traditional, or hybrid project?
NASA uses the traditional waterfall approach to project manage its missions. Data analysis and testing is done on paper or in the computer many, many times so that the chance of real-life success is very, very high. The next mission is a crewed capsule to orbit the moon in 2024. There are no other rocket launches planned before that.
SpaceX, on the other hand, adopts an agile approach to its missions.
Or more precisely, a hybrid project management approach with an overall agile roadmap. Each test launch can be considered a "product release." Each plan, do, check, act (PDCA) iteration is, however, done using a traditional waterfall approach.
Risk is managed by real-world testing, failing, and learning.
As an agile project, there are unknowns. The biggest unknown is how many releases will it take to be achieve the project goal of landing people on the moon--and perhaps, eventually on Mars.
This is akin to the Wright brothers’ approach to achieving first flight in 1903. They expected to crash many times, so they brought tools and materials to repair and fix their gliders on the train journey to Kitty Hawk. In 1902, their third glider design worked and they made 700+ successful flights with it. The year after, they added a small 12hp motor to the glider to achieve powered flight.
They were not supposed to succeed. The smart money, the government, and the press expected Samuel Pierpont Langley to achieve first flight. He started earlier with a successful unpiloted model in 1896. This was followed by two test flights of the full-scale Aerodrome on October 7 and December 8, 1903.
Nine days later, the Wright brothers achieved first flight, on December 17, 1903.
Conclusion
Is there a right way to do project management for space flights? The risks are high, the costs are high, and the possibility of successfully landing humans on Mars is unknown.
In these types of developmental projects where we do not know what we do not know, the most important metric is not a successful launch, but the speed of learning. Every test cycle should produce more data.
In that regard, real world tests produce more accurate data than simulated tests.
Even more importantly, failures from real world tests transform unknown unknowns to known unknowns. Engineers can start working on the problems once they know what they need to fix.
If the success measure of Starship’s test flight was the number of unknown problems discovered, it could honestly be called a smart failure.
Alternatively, if everything had gone well because chance favored the right conditions, some unknown problem could have been left undetected. And this could potentially result in a future explosion. That would not be good.
All said, I am looking forward to Starship’s next test flight to see what progress is made.
This is a guest post by Wan How.
Comments