An ordinary morning flight to Pangkal Pinang had 189 people on board, and the decisions that would kill everyone aboard had already been made, years earlier and far from the cockpit.
Rudi Lumbantoruan managed a plantation in Pangkal Pinang, the town Lion Flight 610 was heading to. His work kept him away from his wife and 2 children for long stretches. The weekend before the flight had held his 11th wedding anniversary. His sister-in-law was about to be married, and he had already used all of his annual leave, so he was not sure he could come back for the wedding. Before he left, he gave her a pair of shoes. “This is a present for you, a pair of shoes, in case I won’t be able to come home for your wedding because I have used all of my annual leave days.”
Adonia Bongkal was flying to Pangkal Pinang with his 2 children, Michelle Vergina and Metiuw Deril, for the funeral of a grandparent. The children’s mother had travelled ahead to help prepare, so she was not on the aircraft. She expected to meet her husband and both children when the flight landed.
Hizkia Jorry Saroinsong was 23. He was flying to Pangkal Pinang to represent AIESEC, the international youth organisation. That morning he sent a message to his family group chat before he boarded.
None of the 3 was doing anything out of the ordinary. Each had put a morning into the hands of a system that asked for nothing except trust. It held for everyone else Lion Air flew that day, but not Flight 610.
Lion Air Flight 610 took off from Jakarta on the morning of 29 October 2018. About 13 minutes later it fell into the Java Sea. All 189 people aboard were killed. The message Jorry sent before he boarded was the last thing his family heard from him.
In most accounts of an accident like this, the number is where the people stop being people, turning into a figure in a report, a line in a chart, the measure of a management failure. A single organisational failure can empty most of a family at once, and a casualty figure hides that.
For those 13 minutes, every one of those lives depended on 2 pilots trying to understand an aircraft that was contradicting itself.
A sensor on the captain’s side, replaced the day before and wrongly calibrated, was feeding the aircraft a false reading that the nose was pitched dangerously high. From the moment of rotation the captain’s control column shook, the stall warning that tells a pilot the wing is about to stop flying, even though the aircraft was climbing normally. The airspeed readings on the 2 sides of the cockpit disagreed with each other. So did the altitude readings.
Once the crew raised the flaps, a piece of flight-control software called MCAS became active. Reading the same false sensor, it concluded the nose was too high and trimmed the aircraft downward. The captain corrected with the electric trim switch and the nose came back. A few seconds later MCAS trimmed down again. The captain corrected again. This went on, a slow contest between a pilot restoring the aircraft and a system removing the correction. No clear message anywhere in the cockpit named MCAS as the source of what was happening.
A procedure could have stopped it. Cutting power to the stabiliser trim would have ended the automatic movement and left the crew to trim by hand. The crew did not complete that step. Near the end the captain handed control to the first officer, the automatic trim ran further than before, and the aircraft went into a descent it did not recover from.
One fact turns this from an accident into a governance story. The same aircraft had done the same thing the evening before. Flying from Denpasar to Jakarta, it had produced the same false stall warning and the same automatic nose-down trim. That crew worked out that the electric trim was the problem, cut the power to it and flew on to land safely. The maintenance record did not fully capture the automatic trim behaviour or the recovery the crew had used, the reported faults were addressed, and the aircraft was released back into service. The next crew inherited the aircraft without the one piece of understanding that had kept the previous flight in the air.
The organisation held the information needed to question whether the aircraft was fit to keep flying. It did not cross the boundary from one decision to the next, and nothing forced it to.
The cockpit was the last place the decision was felt, and it had been forming for years. Boeing built the 737 MAX as a derivative of an aircraft it already made. The commercial case depended on the MAX staying close enough to the older 737 to share a common type rating. The larger, more efficient engines changed how the aircraft handled, so Boeing added MCAS to make it behave like the model pilots already knew. During development MCAS grew far more powerful than first intended, it came to rely on a single sensor, and it was kept out of the manuals and the training. Read on its own, each choice was reasonable. Together, with no one holding the whole, the choices built an aircraft whose safety depended on assumptions that were no longer true.
The choice to treat the MAX as a derivative was made once and then inherited by everyone downstream, and no one was named to keep asking whether it still held. For a decision to stay governable, someone has to own it, its assumptions have to stay visible, the conditions that would make it invalid have to be written down, the decisions that depend on it have to be declared, and someone has to hold the authority to challenge it, pause it or stop it when the evidence turns.
Rudi Lumbantoruan knew none of this. He had boarded an ordinary flight to go back to an ordinary job, carrying a story about a pair of shoes and a wedding he was afraid to miss. The people who carry the consequences of a decision are seldom in the room when it is made, and are almost never told it is being made at all. That is the case for governing decisions while each can still be questioned, before the cost of leaving one alone is paid by people who never saw it.
Kettei: The Discipline of Governed Intent is on Leanpub: https://leanpub.com/kettei-dogi
More on Kettei, the free Boeing case and the Kettei Workshop: https://ishi.co.nz/kettei/
