Forget Maximum Utilization! The Real Magic of Logistics Happens Between the Machines

Motivation – Why Machine Optimisation Alone Is Not Enough

In production and manufacturing logistics, optimisation efforts often focus primarily on machinery and processing equipment. Typical approaches analyse machine performance, failures and efficiency using cause-and-effect diagrams, such as Ishikawa models. This perspective implicitly assumes that improving machine performance will automatically improve the overall system.

However, this view only captures part of reality. Even if a machine is perfectly optimized, it is of little value if it lacks input material, required operating resources, or the ability to pass its output downstream because buffers or transfer points are full. In such situations, machines may be technically efficient, yet the overall system may perform poorly. This highlights a key limitation of machine-centred optimisation: it ignores the systemic importance of material flow and the interdependencies between processes.

To address this, a different approach is needed — one that shifts attention from individual machines to the stability of flow through the system as a whole, and to the buffers before and after them.

The Buffer Model of Logistics: An Alternative View of Production Systems

This model shifts the focus from machines to the buffers that supply and absorb production processes. Rather than asking how well a machine performs in isolation, this approach considers how reliably material flows into a process and how smoothly output can be transferred to the next stage.

From this viewpoint, the buffers located before and after a machine are not merely storage locations, but rather active control elements within the system. An increasing buffer in front of a machine may indicate an overly fast upstream process or a too-slow downstream process. Likewise, a full output buffer indicates that the machine is unable to release its products, regardless of its technical availability or efficiency.

Therefore, the core idea of the buffer model is not to maximise machine utilisation, but to stabilise and control material flow. Provided the inflow and outflow at the buffers are balanced, the system can operate robustly, even in the presence of variability. This logistics-centred perspective provides a basis for viewing production systems as flow systems rather than as collections of isolated machines.

Flow, Buffers and Dependencies: From Chains to Networks

Production systems are often described as chains, but real-world systems are more accurately represented as networks. These systems typically involve multiple input streams, alternative processing paths and shared resources, which creates competition between flows at various points in the system. These interactions significantly increase the complexity compared to a simple linear chain.

The synchronized supply chain in the matchstick game based on Goldratt. For simplicity's sake, we assume that there are always enough matchsticks in stock (100% material availability) and that the customer buys everything we can produce (seller's market)

In such production or supply chain networks, material can flow along different routes, buffers may serve multiple upstream or downstream processes, and bottlenecks can shift depending on demand and routing decisions. Nevertheless, the buffer model of logistics remains applicable. Buffers still function as control elements that absorb variability and regulate flow, even in the presence of competing paths and multiple dependencies.

From a logistics perspective, the objective remains unchanged: to stabilise the flow through the network by managing the buffers rather than optimising individual machines or paths. This makes the buffer model a powerful abstraction for understanding and controlling complex production and supply chain networks.

The Theory of Constraints as a Conceptual Framework

Developed by Eliyahu M. Goldratt in the late 1970s and 1980s, the Theory of Constraints (TOC) was conceived as a response to the limited effectiveness of traditional local optimisation approaches in production and operations management. Goldratt’s core insight was that complex systems should be understood as wholes, whose performance is governed by a small number of limiting factors rather than the average efficiency of all components.

Link zu Amazon:
ISBN 978-0566086656

He formulated and popularised TOC primarily through his well-known book The Goal (German edition: Das Ziel), first published in the mid-1980s. Rather than presenting the theory in a conventional textbook format, the book presents it in the form of a novel. This narrative approach makes the underlying concepts highly accessible, especially for students and practitioners. The book is therefore strongly recommended reading, not only because of its theoretical relevance, but also because it demonstrates that rigorous systems thinking can be communicated clearly, engagingly and memorably.

At the heart of TOC lies the concept of the constraint: the element that limits the throughput of the entire system. Improving non-constraining resources has little impact on overall performance and may even increase instability. Effective optimisation therefore requires identifying the constraint, exploiting it, subordinating all other processes to it and protecting it against variability.

This logic is fully compatible with the buffer model of logistics. Buffers are deliberately used to shield constraints from disruptions and stabilise flow. In this sense, TOC provides the conceptual foundation for understanding why and where buffers are necessary, while the buffer model translates this logic into an operational approach to production and logistics systems.

The Matchstick Game as an Experiment for Dependent Processes

To make the abstract principles of TOC and the buffer model more tangible, Goldratt devised a simple yet effective experiment known as the Matchstick Game. In this simulation, a production system is represented by a sequence of dependent workstations, each of which is subject to random fluctuations in performance. Despite its simplicity, the game reliably reveals counterintuitive system behaviour.

The adapted version used here models a production line comprising five sequential workstations with buffers in between. The potential output of each station varies randomly, while the actual throughput is constrained by the availability of buffers and the capacity of the downstream process. Even when average capacities appear sufficient, the system consistently fails to achieve the expected throughput, thereby illustrating the combined effects of dependency and variability.

As an educational experiment, the matchstick game is especially effective because students can physically execute it. System behaviour becomes observable, measurable and open to discussion. Rather than providing a theoretical explanation of TOC, the experiment enables participants to experience first-hand how local optimisation can fail and why buffers and flow control are essential in real production systems.

The Trade-off Between Work-in-Progress (WIP) and Supply Reliability

A key finding of both the buffer model and the matchstick experiment is that there is a fundamental trade-off between keeping work-in-progress (WIP) inventory low and ensuring high supply reliability. Reducing WIP inventory lowers related costs and increases transparency. However, insufficient buffers expose the system to variability and can quickly lead to starvation or blocking of dependent processes.

This conflict is well known in the literature as the 'materials management dilemma'. Increasing buffers improves the probability that workstations are continuously supplied, but this comes at the cost of higher capital binding and inventory expenses. Reducing buffers improves cost efficiency, but increases the risk of interruptions and throughput losses.

The experiment demonstrates that there is no simple optimum at zero inventory. Rather, system performance depends on a carefully chosen buffer level that stabilises the flow of work in progress (WIP) while keeping it under control. This means that buffer sizing is a strategic rather than an operational decision in production logistics.

Didactic implications and value for education

⧉ Handbuch zum MINT-Praktikum 2017
page 113 ff

From a didactic perspective, combining the buffer model of logistics, the Theory of Constraints and the matchstick game provides an effective approach to teaching systems thinking in production and logistics. Rather than beginning with abstract theory, students can experience system behaviour directly through experimentation, observation and data analysis.

The experiment renders core concepts such as dependency, variability, flow, buffers and constraints visible and measurable. This experiential approach fosters a deeper understanding than purely analytical explanations and aligns well with problem-based and inquiry-driven learning formats, such as laboratory work or simulation-based teaching.

In this sense, the matchstick game serves as a bridge between theory and practice. It provides a simple yet robust experimental environment in which the fundamental principles of production logistics can be explored, questioned and understood — making it particularly well suited to educational settings, and to later transfer to more complex simulations and real-world systems.

The conclusion: From Theory to Action

Theory is merely a compass; practice is the terrain. Don't let this be just another article you read and forget about. To truly understand production logistics, you must experience the system first-hand.

Here is your challenge:

• Go to the shop floor. Don't focus on the machines, but on the piles of materials between them. These buffers reveal the true state of your material flow.
• Play the game: Use the Matchstick Experiment to see how variability reduces throughput. It’s the fastest way to transform an abstract concept into a 'Eureka!' moment.
• Stop chasing 100%. Ask yourself where local 'efficiency' is actually hindering your overall flow.

Don't just optimise — stabilise. Start managing the flow, not just the machines.

Quellen

1. Goldratt, Eliyahu M.; Cox, Jeff: The Goal: A Process of Ongoing Improvement. 3. revidierte Auflage, Gower Publishing, Aldershot 2004. ISBN: 978-0566086656.
2. Wölker M (2017), "Handbuch zum MINT Praktikum" Pirmasens, Germany, October, 2017. [BibTeX] [URL]

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