Flowline

Don't build more — make it flow.

Flowline is a clean idle factory game where the satisfaction isn't building more — it's making the line flow. You find what's choking your factory, feed it, smooth out the chaos, and watch a stuttering, stalling production line turn into a smooth, humming river of goods and cash.

The hook: the bottleneck moves

At any moment, one station limits your whole factory. Feed it, and the line speeds up — then the constraint jumps somewhere new. Chase it, fix it, watch it move. That's the loop.

Secretly a real factory simulation

Under the hood, Flowline runs the actual science of manufacturing operations — Theory of Constraints and the queueing laws of production flow: throughput, work-in-progress, variability, machine uptime (OEE), pull/Kanban scheduling, quality and yield. But there's no math in your face. The equations run the world; you just feel the result.

Want the numbers? Flip on the optional engineer's overlay and every real figure is there — cycle time, WIP, OEE, variability. Most players never need it. Industrial engineers will recognize an old friend.

What you'll do

Find the constraint and feed it — stop wasting money upstream []Buffer the chaos — tame variability with the right amount of WIP, not the most []Keep machines healthy — preventive maintenance vs. run-to-failure []Switch push to pull — discover why just-in-time beats overproduction []Improve forever — prestige as Kaizen: bank permanent, learned best practice []Scale up — connect plants into a supply network where the bottleneck can live in another factory, or in the road between them

Challenge Mode — 18 puzzles, eight chapters

A broken plant, a deadline, and one right answer. Each puzzle is a real diagnosis: capacity, variability, batching, sourcing, market limits, flow time. Every one is proven solvable — and proven resistant to the obvious wrong fix — by an automated suite before it ships.

Checked against the textbook

The simulation's throughput and WIP behaviour is measured against Hopp & Spearman's Best-Case and Practical-Worst-Case bounds in a test suite that runs on every code change. It beats Practical Worst Case by 17–24%, exactly as its own measured variability predicts. If you teach or study operations, the numbers are honest enough to point at.

Idle-friendly: it runs while you're away, with offline progress.

Inspired by Capitalism, Factorio, and The Goal.