VIRTUAL FACTORY

Where the Robot Earns Its Place

Analyze a production line to find the bottleneck and deploy a humanoid work cell where it actually raises throughput, the deployment reality of 2026.

01Challenge

A three-station cell is stuck at 5 parts per minute because Place and Fasten is the bottleneck. Add a parallel humanoid at that station to lift the line to the 8-per-minute target. Automation earns its place only at the constraint.

02Model

The narrative has shifted to deployment, and 2026 shows why: Figure spent eleven months on the BMW line contributing to more than thirty thousand vehicles, Agility's Digit moves totes at Amazon and GXO, Apptronik's Apollo is being tested at Mercedes. But a robot only helps where the line is actually constrained. A production line runs at the speed of its slowest station, the bottleneck, and adding capacity anywhere else changes nothing. You analyze the cell, find the binding station, and put a humanoid there in parallel. This is the discipline behind a good deployment: earn the throughput, do not just add a robot because you can.

03Guided practice

1. Find the bottleneck. Rate per station is machines/time; the line runs at the slowest. Place and Fasten at 12 s/part is the constraint (5/min). 2. Add capacity there. Set Place and Fasten machines to 2 (a parallel human-robot pair), doubling its rate. 3. Re-check. The new bottleneck should clear the 8-per-minute target. 4. Do not over-build. Extra machines elsewhere waste money without raising throughput.

04Feedback

PASS WHEN PASS when the line reaches at least 8 parts per minute by adding capacity at the true bottleneck, Place and Fasten.

If your run fails, check:

Still at 5/min: the bottleneck is Place and Fasten (12 s). Give it a second machine (a parallel humanoid).
Added machines elsewhere and nothing changed: capacity off the bottleneck does not raise throughput.
Over target but expensive: one extra machine at the bottleneck is enough; you do not need more.

05Retrieve & space

What sets a production line's throughput? (the slowest station, the bottleneck).
Why does adding a robot away from the bottleneck not help? (non-binding stations have spare capacity).
Name a real 2026 humanoid deployment and where it runs (e.g. Figure at BMW, Digit at Amazon, Apollo at Mercedes).

06Mastery & project

Find a line's binding constraint and deploy a humanoid exactly there to raise throughput: automation that earns its place.

The deployment reasoning that separates a real productivity gain from a robot for its own sake.

← Agency and Action Efficiency Amplify, Not Replace →

VIRTUAL FACTORY

Where the Robot Earns Its Place

Edit the line config. Put a humanoid where the line is actually constrained and hit the throughput target.

You read

the arrays and values already in scope

You change

the code you write in each cell

Fixed

the dataset and the checks that grade you