Using Factorio as a Logistics Simulation

Why I Built a Factory Logging & Analysis Mod:
Observability in Logistics Education

Factorio is typically viewed as a game centered on efficiency and automation. However, I have always viewed it as something different: a remarkably capable logistics and production simulator. Almost every core concept of operations management, such as inventories, flows, buffers, machines, constraints, and feedback loops, can be modeled inside the game with surprising realism.

link to factorio mod portal

This is precisely why I use Factorio for teaching and experimentation. Not because it is playful, but because it is systematic. For my purposes, what was missing was not another gameplay mechanic, but rather a way to observe system behavior in a structured and reproducible manner. This mod exists to close that gap.

Instead of gameplay automation, the mod focuses on observability.

The mod does not try to make Factorio easier, faster, or more convenient. It doesn't automate decisions, suggest layouts, or optimize blueprints. Instead, it focuses on one specific question: What is actually happening inside the factory over time?

Factorio already exposes a vast amount of information through its UI, statistics screens, and circuit networks. This is not a weakness, but a strength. The difference this mod introduces is persistence and structure, not access to values. By periodically logging inventory changes, machine activity, power usage, and pollution, the factory becomes a time-based system that can be examined outside the game environment.

This fundamental shift in perspective transforms the question from "What do I see right now?" to "How did the system behave over time?" It transforms the player from an operator into an analyst.

Why External Analysis Is the Point

In real-world logistics, analysis rarely happens entirely inside operational dashboards. Decisions are based on data that is exported, transformed, compared, and discussed. Models are built, assumptions are tested, and the system is observed again.

Although Factorio supports building and changing systems exceptionally well, it does not naturally support this analytical loop. This mod enables you to collect clean, reproducible data that can be examined with tools like Excel, Python, R, or advanced monitoring stacks.

The mod intentionally avoids built-in dashboards. In industrial environments, visualization is rarely the bottleneck; rather, it is data quality and structure. I have previously written about how Factorio can be used to explore real logistics concepts, such as push/pull strategies, the bullwhip effect, and pragmatic material flow fixes. This mod provides a direct means to that end. (Reference: Link to earlier article on Factorio and Logistics Concepts).

Education as the Starting Point, Not the Constraint

The original motivation for this mod comes from logistics education. In a traditional classroom, students work with real data, but they cannot alter the underlying physical system to rerun an experiment. In Factorio, however, they can.

They design a factory, collect data, analyze bottlenecks, modify control policies, and immediately see the consequences. Incorrect assumptions become apparent instantly. This makes modeling and abstraction tangible. Importantly, this does not require turning Factorio into a "teaching game"; the game remains unchanged. The mod simply adds observability to an already realistic simulation.

Background: Logic, Observation, and the "Sherlock Method"

Years ago, I wrote a German-language article about logic and reasoning in Factorio, inspired by Sherlock Holmes’ method of deduction. The core idea was that careful observation and systematic thinking are the foundations of technical intuition.

Factorio is an excellent environment for this. Logical circuits and feedback loops are not abstract concepts, but rather concrete mechanisms. Many structures used in automation engineering, such as registers, edge detection, and state-dependent control, can be explored directly. (Reference: Link to German article on Logic and Observation).

This mod is the natural continuation of that idea. If we treat Factorio as a serious logical system, observing its behavior systematically becomes essential.

A Niche Tool, by Design

Some community feedback correctly pointed out that this mod is not for everyone—and that is true. If you prefer intuitive optimization or purely experiential play, you don't need to export data.

This mod is for the experimenters: people interested in system behavior, robustness, and tradeoffs. This includes educators, but also "Megabase" builders who care more about structured measurement than new gameplay features.

Regarding the development process: while I use modern tools and AI to assist with implementation, the design decisions, scope, and purpose remain human. The value lies in what the mod enables the user to do—moving from guesswork to data-driven science.

Current State and Feedback

The mod is usable but intentionally "living." Discussions on Reddit, Discord, and the forums continue to sharpen its positioning. Feedback on metrics and performance is always welcome.

The mod is available on the Factorio Mod Portal: Link to Mod Portal

Factorio does not need to be "gamified" to be educational. It is already a powerful simulation. What it needs—for teaching, experimentation, and serious analysis—is better access to its own behavior. That is what I am building.

Transparency and Augmented Workflow Disclosure

As a logistics professional, I believe in optimizing every process, including communication. In the spirit of total transparency, I would like to share the story of how this project, and even this text, came to be.

The "Augmented" Methodology:

• Human Vision & Logic: The core concepts, educational philosophy (the Theory of Constraints), and logistical system design are entirely my own and are based on years of teaching and consulting experience.
• AI as a Multitool: I treat Large Language Models (LLMs) as a high-speed "technical laboratory." This article, the LogSim mod's Lua code, and the visual assets were developed through a tight human-AI feedback loop.
• The "Skill of Tools": The real work wasn't "letting the AI write," but rather precisely orchestrating tools. This involved knowing where to use AI to bridge language barriers (English/Lua) and where to apply my professional expertise to ensure the data is scientifically sound.

Why I work this way: I don't hide the use of AI; I showcase it. This approach enabled me to transform a complex educational vision into a functional, multilingual tool in a fraction of the time it would traditionally take. By bypassing the technical bottleneck of manual coding and editing, I can focus on my mission of providing students and factory builders with the data they need to understand complex systems.

If you value this tool's insights and data, you are witnessing the result of modern, expert AI collaboration.

Blog: , Seite:
Version: 1.4 April 2025, Kontakt: E-Mail Martin Wölker
Pirmasens, Germany, 2018-, ausgelesen am: , Licence CC BY