Skip to main content
You're offline. Cached data shown.
Lesson 3 of 5

BOM Risk and Second Sources

Read the bill of materials like an underwriter

A flight BOM for a small satellite runs to a few thousand line items. Most are boring. A handful can end the programme. Risk-reading a BOM means finding those lines before the purchase orders go out.

The four questions per line

  1. How many qualified sources exist? One is a risk. One that is also a competitor's subsidiary is a bigger risk. One in a country that could restrict export tomorrow is the biggest.
  2. What is the lead time, and what is its variance? A 40-week part that always ships at 40 is manageable. A 20-week part that sometimes ships at 60 is not.
  3. What happens if it doesn't come? Is there a form-fit-function substitute? Would substituting force a re-qualification, a redesign of a board, or a redesign of the spacecraft?
  4. Is it obsolete or going obsolete? Commercial silicon has a 3–5 year life; spacecraft programmes last longer. Last-time-buy notices arrive at the worst moment.

Single-source exposure

The useful number is not "how many single-sourced parts" but how much of the programme is sitting behind them. Weight each single-sourced line by its replacement lead time and the cost of a redesign if it vanishes. A single-sourced $5 connector with a 6-week second-source qualification is noise. A single-sourced star tracker with no substitute and a 52-week lead is a programme-level risk that belongs on the risk register with a named owner.

Second sources are a design decision, not a purchasing one

You cannot second-source a part after the board is laid out around its footprint and its quirks. Second sourcing has to be engineered in:

  • Design to the interface, not the part. Pick a standard footprint, voltage, and protocol that two vendors satisfy.
  • Qualify both. A second source you have never tested is a hope, not a plan.
  • Split the buy — even 80/20 — so the second vendor's line stays warm and the first vendor knows they are not the only option.
  • Build vs. buy. For some subsystems (structures, harness, simple electronics) the most reliable second source is your own shop.

Obsolescence management

  • Subscribe to vendor product-change notices (PCNs) for every active line.
  • Keep a lifetime buy budget for parts with fewer than five years of expected availability.
  • For processors and FPGAs, prefer families with a public long-term-supply commitment (the space vendors publish these; commercial ones rarely do).

Score your BOM

The calculator below produces a simple exposure index: single-sourced lines weighted by lead time and redesign cost, relative to the whole programme. It is crude by design — the value is in arguing about the inputs with your engineers.

Calculator

Single-Source Exposure Index

Rough programme-level exposure. Redesign cost is what it would take to engineer around a vanished part.

Share of critical lines single-sourced
100 * single / lines
2e+1%
Expected annual redesign cost
single * (pfail/100) * redesign
error$k
Expected annual loss vs budget
100 * single * (pfail/100) * redesign / (budget*1000)
1.9%
Exposure index (0 = none, 100 = severe)
min(100, (100 * single / lines) * (lead / 52) * (1 + (single * (pfail/100) * redesign) / (budget*1000)))
error
Finished this lesson?
Progress is saved in this browser; sign in to keep it across devices.