Systems & Hidden Structure

The Factory as a Living Network

Personal interest
Observed in experiments
Personal metaphor
Working interpretation

Building aircraft brings together thousands of tasks, people, tools and parts that depend on one another, while priorities keep changing. Working in that setting taught me how making one area efficient can clash with keeping the whole flow moving.

What happens across the whole system when one task runs late?

Why it attracts me

Aircraft production combines thousands of tasks, people, tools and parts, and the priorities keep changing. Working in that setting, I came to see a factory less as a list of jobs and more as a web, where each task waits on others and each finished task releases more.

The idea

In a web like this, local efficiency and overall flow can conflict. A team can hit its own targets by working on what is easiest for it, while the work that others are waiting for sits untouched. Delays spread along the links, often far from where they began. To see what is really happening, I need the network itself (Networks: The Math of Connections), not just a scorecard for each area. It also means seeking the true limit on flow (Finding the Real Bottleneck), not rewarding busyness.

An example

The Availability Architecture Lab on this site shows a small version of this. Three machines work in a row, and the whole line stops whenever any one of them breaks. With the starting settings each machine is up about 97% of the time, yet the whole line is up only about 91%. Adding a backup machine helps, but less than expected when one shared cause, like a power dip, can stop both.

Where it connects

The same pull between the part and the whole shows up outside work. The drive to improve things on the job can compete with the limited attention I want to give at home (Being There Versus Achieving More). And the people inside the network matter as much as its structure (Engineering That Respects How People Work).

What this does not establish

Calling a factory a living network is a picture that helps me think. It does not mean a factory is alive or behaves like a living body.

Questions I'm still exploring

How can a tool show the ripple of one late task before it reaches the end of the line?
Which local measures quietly work against the flow of the whole?
How do I respect each team's skill while asking them to serve the flow of the whole?

Sources and further reading

Wallace J. Hopp and Mark L. Spearman, Factory Physics (McGraw-Hill)

Working interpretation: drafted from my notes and interests for review. It is not a direct quotation, and I may still change it.

In short

A large factory is less like a single machine and more like a web. Each task waits on parts, tools, people and other tasks, and each finished task frees up others. A delay in one place spreads along these links. Because of that, a team can look very efficient on its own numbers while the factory as a whole moves more slowly. Calling the network living is a picture: it keeps changing as priorities, people and problems change.

Why this matters to me

Aircraft production taught me a lesson I keep returning to: local efficiency and overall flow are different goals, and they can pull against each other. That shapes what I want from the tools I build. I want them to show how a delay travels through the web of dependent work, not just which area missed its own target.

Connections

Being There Versus Achieving MoreTension

The drive to improve things at work can compete with the limited attention I want to give at home.

Seeing Connections, Not Just PartsPart of

Aircraft production is a concrete case of systems thinking: thousands of linked tasks, where making one area efficient can slow the flow of the whole.