THE HIDDEN WORLD OF MANUFACTURING
We often discuss digital technology. Some of us build digital identities, others move their businesses online, and workplaces undergo digital transformation…
Today, “digital” dominates the language of the modern world, but how much of our lives does it really shape, and in what way?
In my view, not much, yet it is indispensable, like salt. After all, whatever we manage to digitalize tends to become more efficient, more current, more competitive, doesn’t it?
And yet, the products that define modern life can only come into existence through conventional means: the fabric of our clothes, the yarn spun somewhere, the ceramic tiles that came out of a kiln, the glasses that were once sand. This is the real thing, physical manufacturing: phones, headphones, remote controls, door handles, chairs, bags, keys, chocolate, wafers… Whatever comes to mind. All of them planned, all of them produced through certain processes, all of them made possible by the work of many people. Are we really aware of this?
I do not think so, as our conversation rarely moves beyond being online and working remotely. Tools, in many ways, have turned into ends. We keep talking about the service sector, yet if nothing has been produced, what exactly are we going to serve? Take tourism, for example. Success in tourism requires not only an excellent construction sector, but also the ability to offer products and facilities that can host visitors at the highest level of comfort. If you do not have an industrial base to support this, then at the very least, you must have the wealth to source it from elsewhere.
Still, we only notice this reality when something goes missing. Think back to the pandemic, those days when shelves were empty, and we could not find or access what we needed. Or the crisis in the Suez Canal in 2021, when a giant container ship ran aground due to a sandstorm and poor weather conditions, disrupting shipments.
The daily loss to global trade reached 10 billion dollars. And this was only a transportation crisis. Imagine if it had occurred at the stage of production or harvest.
How Things Are Made: A Journey Through the Hidden World of Manufacturing was written to create precisely this awareness.
Its author, Tim Minshall, recalls first seeing this awareness in the curiosity of children in a classroom. Invited to a school’s science week while he was an academic, he brought computers, keyboards, cables, and circuits. The children examined the parts and asked questions. The labels read “Made in Malaysia,” “Origin: Mexico,” and “Assembled in the USA.” Manufacturing, it turned out, was spread across the world.
There are also historical and social dimensions to this. In the past, we knew the producer: the tailor, the butcher, the baker. Today, production is fragmented, and supply chains are long and complex. Yet we simply click to place an order, it arrives at our door, and everything seems to work flawlessly.
Tim Minshall is the Dr. John C. Taylor Professor of Innovation at the University of Cambridge and heads the Institute for Manufacturing within the engineering faculty.
The book examines the stages of how manufacturing actually works.
REMEMBER: IN THE REAL WORLD, YOU CANNOT TASTE A CHOCOLATE WAFER THAT HAS NOT BEEN MANUFACTURED.
Even behind the simplest things, there is an entire world at work. To really understand this, think about what lies behind an everyday product, a product whose technology barely even registers with us. Take toilet paper, for example. Hundreds of small decisions, years of trials, careful calculations. Each sheet has a weak line along its edge so it can tear easily. Multiple layers cling to one another to create volume. The texture is soft, sometimes cushioned, and it dissolves in water. The story of toilet paper manufacturing begins in a forest, where trees are suitable as raw material. A former partner of mine, one of the leading players in Europe’s hygiene paper sector, owned private forests covering an area comparable to Belgium. The trees were carefully prepared and converted into packages of pure cellulose. Bark, sawdust, chips, and similar residues were diverted to other manufacturing industries. Paper pulp was then prepared according to a specific formula and processed with great precision in enormous, high-technology machines. At one point in my career, I was a leading player in Türkiye’s packaging industry, with factories processing various types of paper. It is a sector where every stage and every detail is meticulously designed, production runs without interruption, and energy consumption is inherently high. Even today, I hesitate whenever I use hygiene paper. I think of those forests, the scale of capital investment, the energy consumed, and ultimately, nature that dissolves in water and disappears, never to be recovered. When I was a child, in certain conservative circles, it was even considered religiously inappropriate to use toilet paper. Yet I also remember my mother’s laundry days. When I was little, she would probably have me wash used handkerchiefs simply to keep me occupied. So, what can we do? We use hygiene paper while it is available. I no longer manufacture it myself.
The packaged toilet paper we see on the shelf and buy is the result of a finely coordinated production system that tracks the movement of thousands of product types, from factories to warehouses and on to stores. When demand and supply fall out of balance, chaos follows, and the system must be rebuilt.
In fresh food, the balance between supply and demand is far more delicate. The soil is prepared, seeds are planted, crops are grown, and harvested. The product then enters the cold chain, is sorted at distribution centers, and directed to stores. This is a race against time. Even small changes in weather conditions affect both demand and yield. When seasonal labor is insufficient, crops remain in the field. When one link in transportation fails, the finished product does not reach the shelf on time. What looks like a simple chain is, in fact, long, fragile, and costly.
As the Scale of Production Grows, Production Networks Expand. As Networks Expand, Fragility Increases.
As scale grows, dependencies grow as well. A car consists of thousands of parts, an aircraft of millions. These networks cannot be described with straight lines. As Minshall points out, where the network becomes unclear, risk becomes unknown. This is why we, as manufacturers, invest in data analytics, network mapping, and AI-supported monitoring systems. Today, people on the production floor process this information online, since even a small disruption can quickly spread across the entire system.
The balance of the supply chain is rebuilt every day. Each stage and each input operate on its own timeline, from harvest cycles to festivals and holiday seasons. Historical data helps anticipate seasons and peak periods, but every forecast carries a margin of error. Unexpected events only widen that gap. Being prepared for every possible scenario is neither realistic nor feasible, given limited resources. This leaves manufacturing with a fundamental question: how to maintain efficiency and resilience at the same time.
Minshall argues that we must first learn how to see, and only then learn how to ask the right questions. Where is the system fragile? Where is there room to absorb pressure? Which decision frameworks strengthen it? Answering these questions properly requires the right perspective.
What Happens Inside a Factory?
Inside a factory, everything can be understood through seven core elements: inputs, methods, people, machines, materials, process, and output. The product may change, but the cycle largely remains the same. When each step in the process is well designed and runs smoothly, quality production is the result.
As Volume Increases, Organizational Structure Changes.
At the most fundamental level, production can be organized around four distinct structures. Workshops and one-off jobs prioritize flexibility. Batch production delivers large quantities of the same product. In mass production, standardization is essential. In continuous flow, the product is singular and produced at scale.
All of these production models, however, require knowledge, decisiveness, agility, and adaptability. Management’s experience and intuition, in turn, provide a clear advantage. One of the roles of automation, when designed around accumulated experience and know-how, is to capture that intuition, make it measurable, and turn it into early warning signals.
The diversity we see in production today has historical roots. The critical step in the shift from craftsmanship to industry was the concept of standardized, interchangeable parts. This made it possible to standardize measurements and produce each component within defined tolerance ranges, allowing parts to replace one another. The assembly line followed, introducing speed and repeatability into production. Quality was no longer something inspected at the end; it became embedded in the production process itself. Data collected from the shop floor and measurements taken from machines then became part of planning. Production increasingly came to rely on the flow of skills and knowledge.
As for distribution, the large trucks and lorries we see on the roads are also integral links in the production chain. Some carry raw materials, others semi-finished goods, and others transport packaged products to distribution centers. In manufacturing terms, this movement is known as “logistics”. At its core, logistics is about delivering the right product to the right place at the right time. And just as the flow inside a factory depends on planning, logistics requires the same discipline and a well-designed plan.
As products become more complex, the network through which they move expands as well, and the entire process rests on precise planning. Success depends on getting the right parts together, on time and in the right order. When this coordination works flawlessly, no one notices. When something goes wrong, however, the entire picture becomes visible at once and, depending on its scale, quickly turns into a topic of discussion.
You might ask whether all this effort is really necessary. Is there not an easier way? In fact, this is the most efficient one. The best version of any component comes from its strongest producer. What follows is simply a matter of cost and capacity. Production has also spread across geographies for a simple reason: standards made transportation cheaper. Put goods into large metal containers, load them onto ships, then onto trains, and finally onto trucks. In a system where ports, vessels, handling equipment, and software all speak the same language, transportation became both faster and cheaper. As a result, production spread across the world and could operate as a single, interconnected system.
Looking at the other side of the coin, a different picture emerges. Certain economic calculations can make it appear rational to move a product thousands of kilometers. Decisions that look sound on paper often leave environmental impacts out of the equation. The emissions, noise, pressure around ports, and the effects on marine and terrestrial ecosystems created by these operations are rarely fully accounted for. Yet the core principle of production is to reduce waste, and hidden costs are part of that waste. At this point, two priorities stand out: transport less and transport more cleanly.
Turning to the human dimension, none of these processes function without people. Operators at ports, planners behind screens, drivers on the roads, and producers on the factory floor all strive to do their best. Since the margin of error can never be reduced to zero, small deviations are corrected every day. The supply chain finds room and resources to recover. Of course, there are breakdowns and returns. Repairs are made, data is updated, and plans are revised.
Reading demand is a complex task. When order quantities and delivery dates are clear, manufacturers can assess capacity, suppliers, and pricing and make decisions. But most of the time, this is not the case. Customers often do not know exactly what they want, or when they want it. Managing this uncertainty is our responsibility.
After the Industrial Revolution, the era of made-to-order, fully customized production came to an end. Manufacturing shifted to a make-first, sell-later logic. Production no longer depended on the immediate existence of demand; supply itself could generate demand. Over time, companies developed expertise in shaping and steering demand. Forecasting demand, however, remains difficult, as the members of the supply chain and consumers often have competing expectations and interests. This is where management skill truly matters: establishing balance to serve all stakeholders. Volume, variety, and variability sit at the center of this equation. Without properly understanding these three, nothing from line design to material planning can operate at full efficiency within the system.
Forecasts, of course, do not always come true. More often than not, they are imperfect. Historical data helps guide decisions, but it is rarely sufficient. At times, a product addresses an entirely new need, and customers cannot immediately grasp what it is. For this reason, manufacturers must be able to view situations from different perspectives. Scenario planning is conducted, field insights from sales teams are gathered, data from digital dashboards is reviewed, and a shared view is formed. Yet in the end, a forecast is still an assumption, albeit one informed by multiple sources of data. At a fundamental level, there are two production approaches. The first aims to keep production as stable as possible throughout the year and manage fluctuations through inventory. This approach works well for products with long shelf lives. The second closely tracks demand and adjusts production to match it. This requires speed, reliable forecasting, and a supply chain capable of responding just as quickly. When contingency plans are in place, sourcing alternatives exist, and teams are experienced enough to step in at the right moment and in the right order, balance can be maintained.
Digitalization makes things easier, but not always! With internet and mobile data access, retail systems have made flows visible on a daily, even hourly basis. Data volumes have multiplied. At the same time, consumer expectations of manufacturers have risen. We have come to see constant, everywhere availability as the norm. In an ever-changing world, the manufacturer’s task is to meet demand as efficiently as possible by finding a workable balance between new expectations and the system’s real capacity.
The manufacturing industry cannot succeed without government support! Large-scale transformations, such as the automotive industry’s shift to electric vehicles, are not challenges companies can tackle on their own. Governments support these efforts through a range of instruments, including financing and operational support. Minshall defines the role of the state across three core areas:
• Supply side: Mandating or supporting technological transitions. The United Kingdom’s decision to ban the sale of gasoline and diesel vehicles from 2030 onward is currently on the agenda.
• Demand side: Encouraging the adoption of new technologies. Subsidies for electric vehicle purchases and investments in charging infrastructure are a clear example.
• Managing the impacts of transition: Addressing job losses, skill shifts, and the broader social effects created by these transformations.
The new R&D laboratories of manufacturing are the research ecosystems within universities that work in direct contact with industry. In these environments, manufacturers test new technologies, engineers work alongside student’s redesign processes, and capabilities such as automation, robotics, and data collection are developed. Another important value these centers create is the removal of scale barriers. A large automotive manufacturer and a local workshop can test on the same production line. Small and medium-sized enterprises can experiment with robotic arms, sensor systems, or digital twin software that would normally be beyond their reach.
Minshall also reminds us that not every transformation is revolutionary. Many innovations advance through small, incremental steps. The bulky mobile phones of the 1980s have evolved into palm-sized devices used by billions today. This is a classic example of incremental innovation. Products may become smaller, more powerful, or more affordable over time. This is precisely why the manufacturing world never stands still. Customers adapt to products, and companies reflect, refine, and improve. This cycle is inherent to manufacturing.
Change alone is not enough. Unless new components fall into place, data connects meaningfully, and processes move in the same rhythm, neither speed nor quality can be sustained. In the picture Minshall draws, the next step is bringing knowledge and product together on the same line: connectivity.
Today, the scale of connectivity has expanded. The internet first connected research networks, then commerce, and eventually the devices in our pockets. Machines now listen, measure, and share what they learn. Industrial IoT collects data from everywhere, from motors and conveyors to automated guided vehicles and energy lines. Digital twins built on this data make it possible to test scenarios on screen before applying them in the field. A parameter is adjusted first in the twin, without risk. If it works, it is then implemented on the line. This reduces the cost of “learning through error” and accelerates the path to the right outcome.
Another cornerstone of connectivity is integration. The production cycle progresses through four stages: sense, decide, act, and operate. The robot is merely the tool of the “act” stage; sensors, PLCs, and software form the backbone that binds these stages together. Fragmented automation, when not integrated, does not generate efficiency and may even conceal problems. The real task is to gather data into a single pool and manage the flow as one. When that happens, quality becomes visible in real time, bottlenecks are identified at their source, and maintenance is scheduled before failures occur.
Here, Minshall offers four important warnings.
First warning: The future does not offer equality. Today, the vast majority of manufacturing companies are small and medium-sized enterprises, and many still rely on Excel spreadsheets, paper forms, and manual tracking systems. In a world where digital transformation dominates conversation, thousands of manufacturers still struggle to collect their data, let alone interpret it.
Second warning: Digital convenience comes at a high cost. Behind one-click delivery stands people working in shifts, low-margin systems, and fragile supply chains. Order fulfillment depends on the rhythm of many invisible hands. Digital convenience does not eliminate physical labor; when we talk about efficiency, we must not lose sight of human effort.
Third warning: The speed of automated decision-making brings its own challenges. Artificial intelligence systems can now reveal hidden dependencies, bottlenecks, and delay risks across supply chains. Yet if the logic behind these decisions is not transparent, manufacturers cannot understand what has changed or why. This is why Minshall stresses the importance of automation being traceable and interpretable. If you cannot trace a decision, you cannot correct an error.
Fourth warning: Most critically, the entire digital world rests on a physical foundation. Manufacturing now depends on data, and data depends on infrastructure. On one side lies the backbone of the internet; yet without energy, meaning electricity and cooling, and without secure connectivity, no data flows. On the other side stands the semiconductor ecosystem, a system operating with extreme precision that enables billions of flawless chips to be produced repeatedly. If either of these foundations falter, the digital world can no longer carry the load it has taken on.
When we ask ourselves what we can and should do, the pandemic inevitably comes to mind: a single crisis teaches more than a thousand pieces of advice. After the crisis, the distance between factory and consumption shortened, and the line between producer and user blurred. Pharmaceuticals offer the clearest example. Under normal conditions, it can take a decade for a vaccine prototype to move from the laboratory to the patient. During the pandemic, a single, urgent objective compressed the entire timeline. Risk capital and public funding reduced waiting periods, and several steps progressed in parallel. By the end of 2020, vaccines had been approved, and millions of doses had entered production lines. We realized that the threshold we once believed impossible to cross could, under the right conditions, indeed be crossed.
Creative solutions also emerged in logistics. Borrowing another company’s distribution network proved, on the ground, that synergy can work. On a different front, BioNTech’s quality-certified modular production units, designed to fit inside containers, made it possible to move the factory to where it was needed. In pharmaceuticals requiring a cold chain, or in situations where large volumes are needed quickly, these localized solutions provide a clear advantage.
We also witnessed how much risk comes with reliance on distant, fragile, single-channel supply chains. Local capabilities, modular production, and flexibility supported by digital connectivity strengthen both quality and access, in times of crisis as well as under normal conditions.
As manufacturing transforms and accelerates in this way, how will we protect the environment, resilience, and social justice at the same time? Energy use, waste, local employment, and data ethics. All of these must be part of the core plan. In short, we need to discuss a production model that operates in harmony with the limits of the planet and the life it sustains.
So, what can we do, and what should we do in manufacturing? Shelter, food, and clothing are our most basic needs. The production systems we have built around these three make life easier in the short term, yet over time, they have evolved into systems that make life harder. Cities and roads expanding through cement production are simultaneously driving excessive global emissions. Food waste stretches from field to shelf. In clothing, fashion encourages waste through short-term use. As it stands, our current system is not sustainable!
The first step may be to produce with less harm, and to produce less harmful things. Minshall offers concrete examples: a zero-waste approach in industry, where every by-product is turned into a valuable input, and the application of good manufacturing principles that reduce environmental impact while remaining aligned with sound business logic.
The second step is to rethink transportation. Move less. Move cleaner. Delivering construction components to sites as ready-made, large units and turning construction sites into assembly areas reduces truck traffic and waste. Localized food solutions, including urban vertical farming or production closer to consumption, as we witnessed during the health crisis, follow the same logic.
The third step is to reduce consumption intelligently. Instead of a use-and-discard mindset, Minshall describes a circular approach in which reduction, reuse, and recycling operate together. This includes options ranging from repair to rental to service models such as pay-as-you-use. These models encourage manufacturers to design products that are durable and efficient, while allowing users to consume only what they actually need. Going one step further, additional loops such as “refurbishing”, “repurposing”, and “remanufacturing” come into play.
If what needs to be done is clear, why are we not moving faster to fix everything? On this question, I do not fully agree with Minshall. He approaches the answer along three axes. The first is ownership and responsibility. Measuring direct emissions is relatively straightforward, but managing the impacts buried deep within supply networks, extending all the way to the point of customer use, is far more difficult. The second is balance. Decisions that are already hard to make between quality, speed, and cost must now also incorporate sustainability. Even a single product brings dozens of variables into the same equation, from cotton and dye to water treatment and logistics. The third is inertia. Transforming into a large and complex system requires coordination. The intent of a single factory is not enough; adjacent industries must move in sync.
As I said, I do not fully agree. Minshall devotes pages to the subject and offers a long list of recommendations. In my view, many of them are misplaced and unnecessary. They resemble solving a high-school physics problem under “normal conditions,” detached from the realities of industry. You cannot impose environmentally responsible behavior, waste avoidance, or conscious consumption through regulation or financial incentives if these habits were not learned early in life. In fact, these principles should already be observed if the goal is to run a business properly, soundly, and profitably.
Take our own experience. In our distribution operations to retail outlets, we implemented measures to prevent waste and, in doing so, eliminated thousands of vehicle trips and saved hundreds of thousands of kilometers. Naturally, this resulted in lower carbon emissions. Similarly, at Şok Marketler, digitalization and the use of artificial intelligence enabled clearer communication and stronger control across operations, allowing us to achieve faster results.
By contrast, there was a home-delivery business model, name unnecessary, that promised free and immediate delivery of whatever you wanted, in whatever quantity you wanted. I saw this as pure wastefulness. The global success it achieved in a short time, unfortunately, proved unsustainable.
Looking ahead, industrial policies must be designed with an understanding that is aligned with megatrends, flexible, inclusive, and respectful of ecological limits. Without falling for post-industrial society narratives, we must keep manufacturing strategically at the center.
High value-added production areas such as defense, automotive, food technologies, electric vehicles, machinery, and biotechnology will define Türkiye’s growth potential. Yet the inputs of production are no longer limited to steel and concrete. Data, design, code, and green energy are now equally critical. Automation, artificial intelligence, and robotics do not eliminate production; on the contrary, they reinforce it and redefine the relationship between production and labor. For countries like Türkiye, this makes it essential to establish a new balance between protecting the workforce and increasing productivity.
For example, digitalization efforts in organized industrial zones should go beyond machinery alone. They must also be supported by equipping employees with new skills. This requires a shift toward a policy of smart manufacturing combined with smart people.
Global value chains present both opportunities and a potential trap for developing economies. Türkiye often occupies the role of an assembly or intermediate production link within these chains. Yet future advantage in manufacturing will belong to countries that also control design and brand ownership. Expanding Türkiye’s branded manufacturing strength through examples such as Togg, Baykar, Beko, Vestel, Ülker, and Godiva is therefore strategically important. State industrial policies and incentives should aim not only to boost exports, but also to build intellectual property ownership, patents, copyrights, and know-how.
In Türkiye, industrial policy should focus not only on investment incentives but also on institutional capacity, workforce training, and reducing regional disparities. In other words, instead of traditional production subsidies, we need to think in terms of ecosystem support: technoparks, university-industry collaboration, start-up accelerators, and green transformation centers, all scaled through measurable outcomes. Without question, Türkiye has the potential to become a strong central industrial actor in the 21st century. I know this, I write this, and I say this. But it will require sustained effort.
Note: This open-source article does not require copyright and can be quoted by citing the author.
