Proof Defence · embedded electronics

See the phenomenon before fixing the design

≈ 3 weeks · the real cause, not blind reinforcement

The problem

A device failing its drop tests.

During drop testing, an embedded electronic device suffered a mechanical problem involving the housing and the printed circuit board (PCB) — which classic design approaches couldn't resolve properly.

The governing constraint

We were fixing blind.

The real behaviour during impact was too fast and too complex to be inferred correctly from the device's final state. Without seeing the phenomenon, every fix was a bet.

The approach

Change the means of observation first.

Before touching the design, a high-speed camera to see what actually happens during the impact — identifying the resonance and the mechanical behaviour at play.

Then act on the cause, rather than arbitrarily reinforcing the whole assembly.

The result

≈ 3 weeks
problem solved, where classic approaches failed
1 cause
treated at the source, without over-engineering the device

Defence sector — strong anonymisation: no name, no date, no identifying detail. Only the technical problem, the method and the result.