Why Your Engineering Decisions Vanish Into Thin Air

The hidden failure mode that UPS Flight 2976 exposes

The UPS MD-11 engine separation on Flight 2976 sparked the usual post-incident analysis. What happened? Who failed? What regulation got bent?

Those questions matter. But experienced engineers know the real story rarely starts where the story becomes visible. Failures are endpoints, not beginnings. The engine didn't just decide to fall off mid-flight. Somewhere in the years before that flight, decisions accumulated. Trade-offs were made. Known limitations got worked around. Institutional knowledge walked out the door with a retiring engineer.

This is the problem Andre Wegner, CEO of Authentise, and Thomas Rees, Innovation Lead at ToffeeX, brought to RAPID this year. Call it lost engineering intent.

"We built systems to record engineering outputs," Wegner said. "What we didn't really build were systems that preserve engineering thinking. The rationale, assumptions, and trade-offs that explain why something happened in the first place."

"Why was one design path rejected? Why did an experienced engineer decide the technically elegant solution was not worth the manufacturing risk? Those decisions frequently live in design reviews, supplier conversations, and institutional memory. Then institutional memory retires."

The gap between what we save and what we know

Here is what most engineering organizations do well: requirements get documented. CAD models sit in version control. Simulation results get stored. Compliance frameworks exist. Change requests, maintenance records, and validation reports prove work happened.

Here is what most organizations do poorly: the reasoning behind those outputs disappears.

Why was a particular tolerance chosen over another? What manufacturing risk prompted the team to accept a heavier component? Which certification pathway got rejected because of documented reliability concerns that never made it into the formal record?

These questions live in whiteboards, Teams chats, hallway conversations, and the heads of senior engineers. Then those engineers retire, or join competitors, or simply forget. The old system worked when teams sat together, when expertise transferred informally over lunch, when somebody nearby always remembered why a decision got made five years ago.

It was messy. It was occasionally chaotic. But it worked.

Why this is getting worse, not better

Engineering has changed in ways that expose this gap:

Product complexity has exploded. A modern aerospace component might involve twelve suppliers across four countries, each with specialized knowledge that no single engineer holds.

Certification requirements have multiplied. FAA, EASA, AS9100D, NADCAP. Each audit asks for evidence that decisions were made correctly. Few ask for evidence that reasoning was preserved.

Distributed teams have become standard. The engineer who needs to understand why a 2019 decision still constrains 2024 design might sit in a different timezone, never met the original decision-maker, and has access to a report that says what was decided, not why.

Workforce mobility has accelerated. Organizations are losing decades of tacit experience faster than they can transfer it. The average aerospace engineer stays at a company for 4.2 years, according to some estimates.

"Smart people remembering things" stopped scaling. The informal knowledge transfer system that kept engineering intent alive for decades is breaking down faster than most organizations realize.

What gets lost costs money

Missing context mostly slowed organizations down. Teams repeated mistakes. Engineers revisited decisions already made. Occasionally, someone rediscovered a lesson the organization had learned and forgotten three times already.

This is expensive. Estimates for rework in complex manufacturing range from 15-25% of total engineering labor cost. Some of that rework stems from missing context. Someone didn't know why a particular approach got rejected, so they tried it again.

In high-conformance industries, missing intent creates legal and regulatory exposure. Auditors can prove a requirement was met. Proving that the reasoning behind a design choice remains defensible five years later is harder when nobody remembers that reasoning.

"Several of the most important design decisions in an engineering project are never written down," Wegner and Rees noted during their RAPID session.

This is the bottleneck they identified: organizations have invested heavily in digital engineering tools that capture outputs. They have invested far less in systems that capture reasoning.

The path forward

Wegner and Rees are building toward something different at Authentise and ToffeeX. Rather than treating documentation as a compliance burden, they frame it as engineering infrastructure. Capturing intent should be as automatic as version control, as searchable as a requirements database.

Whether their specific approach gains traction remains to be seen. But the problem they identified is real and getting more acute. As engineering complexity increases and distributed work becomes permanent, the cost of lost intent will keep climbing.

Organizations that figure out how to preserve engineering reasoning, not just engineering outputs, will have a genuine competitive advantage. Those that don't will keep rediscovering lessons they should have learned the first time.

M4S TAKE

My take: certifications like this matter because they give buyers a defensible reason to shortlist a supplier. In a market where everyone claims quality, third-party validation is the difference between being considered and being ignored.

Simon McLoughlin