Designing for Failure: Lessons from Metal, Code, and Memory
✈️ Mechanical Case: The DC‑10/MD‑11 Pylon
- The DC‑10 and MD‑11 mounted their massive engines on pylons attached by just three lugs and a spherical bearing.
- This concentrated load path made the system vulnerable: one fatigue crack could mean catastrophic separation.
- American Airlines Flight 191 (1979) and UPS Flight 2976 (2025) both proved the danger.
- Other aircraft, like the 747 or A330, spread loads across multiple fittings, ensuring redundancy.
- The difference is stark: one design assumed failure, the other assumed efficiency.
💻 Software Case: Timers Instead of Sensors
- Modern systems often replace mechanical switches with software timers.
- Example: gear doors commanded to open for 11 seconds, because the process “should” take 8.
- Without a proximity sensor, the system assumes success.
- If hydraulics lag or a door jams, the next command collides with closed doors — unsafe operation born of efficiency.
- Here too, the philosophy shifts from verified redundancy to assumed sufficiency.
🛠 Personal Case: The Home-built VW Aircraft
- I once inspected a home-built plane with a VW engine, certified for aerobatics.
- The paperwork inspired confidence, but the wing brackets told another story: thin aluminum, long bolt holes, minimal strength.
- My industrial mechanical background screamed caution.
- To me, those brackets should have been built to withstand a nuclear explosion.
- An aircraft isn’t an aircraft if a wing attachment fails. It becomes a brick.
- The certificate assumed efficiency; my instinct demanded resilience.
⚖️ The Restorationist Lesson
Across these cases, the pattern repeats:
- Efficiency trims away redundancy.
- Assumptions replace verification.
- Confidence replaces scars.
The old guard was designed for failure because they remembered failure. The new guard designs for efficiency because they trust models. True stewardship requires both the wisdom of scars and the power of models. Restoration means reclaiming the philosophy that safety lies not in efficiency alone, but in resilience — in designing systems that can absorb failure without collapsing.
Conclusion: Whether in pylons, software, or home-built brackets, the principle is the same: assume failure, design for survival. Efficiency may win applause, but resilience saves lives. The restorationist path is to honor the scars of history and insist that every design — mechanical or digital — carries the strength to fail gracefully.