What Software Engineering Should Learn from Aviation

12 Jul 2025 #coding

Flying is safe but… accidents happen.

When they happen, it’s all over the news. But the thousands upon thousands of safe flights don’t make it to the headlines.

The magic of the Internet took me to Admiral Cloudberg on Medium. Each post breaks down a rare accident, tracing the failed part or procedure that triggered the disaster.

After every accident, there’s an investigation

After binge-reading some of Admiral Cloudberg’s deconstructions, what struck me was the meticulous investigation after every accident.

A committee finds out exactly what happened and why. Their task is to find the root cause and the subsequent chain of events.

After finding the root cause, they:

• release bulletins to manufacturers
• update procedures and checklists
• add the accident scenario to simulators to train pilots

Their goal isn’t to blame the captain or anyone else, but to prevent the same mistakes in future flights.

Coding isn’t flying

As software engineers, often we hear or say,

• “Something went wrong. We don’t know why. It hasn’t happened since then.”
• “OK, let’s move on. If that happens again, we’ll take a deeper look.”

Imagine if we treated coding the same way:

• Checklists to release code
• Another checklist for incidents
• Clear onboarding processes
• Automated best practices
• Guilty-free culture

Coding would be as safe and reliable as flying.

We must adopt the same commitment to safety protocols and procedures to never let the same mistake happen twice.