Logistics is the art of finding order in chaos: without standards and data, technology fails

The ideal of fully automated logistics — is it actually achievable in practice?

Logistics is a dual story. On one hand it is pure engineering; on the other, it is a social science. It involves a vast number of technical solutions, yet it constantly demands connection and integration with the broader community. The fact is that things simply do not work without people. In engineering, 100% reliability does not exist — we operate at perhaps 99.9% — and the same holds true for logistics.

The ideal you describe is correct as a guiding principle, but it is not fully achievable in practice. The boundaries keep shifting, and something new always emerges that enables further improvement. We will never standardise everything completely, because nature itself continuously evolves and establishes new equilibria. At its core, logistics is the art of finding new, efficient solutions. But order must exist — without it, chaos follows immediately.

Where do you see the biggest bottleneck in ensuring traceability from the production line to the retail shelf?

The two main challenges are digitalisation and identification. Traceability of shipments is in principle ensured, since goods carry their own serial identifiers. The real challenge is tracking and monitoring accompanying data in real time. We need to ask ourselves what is actually worth tracking. Imagine a truck carrying pork carcasses that overturns, and everyone involved waits hours for an inspector. If you must maintain an unbroken cold chain for that type of cargo, traceability becomes extremely demanding. Will that meat be processed into pâté — and will the pâté be properly labelled with its origin? Products like that should carry temperature-sensitive labels, and the system should know how to respond to deviations. In reality, if we strictly applied every theoretical ideal, we would have to discard a significant percentage of "fresh" products every single day. That is neither practical nor sensible, even if standards must be set high.

The bottleneck, therefore, is whether existing identification systems are capable of capturing and managing that data. A smart "product ID" should contain far more logistics data and be directly linked to sensor technology. We need to solve two things: decide meaningfully what we actually need in real time — which is logistically extremely difficult — and determine the frequency of data sampling. Then we need to connect those applications to each other. Development is moving forward: we started with product codes, progressed to barcodes, today we have QR codes, and the question is when optical condition recognition will become the standard. Today, the physical flow of goods works well; the information flow lags behind — key data such as precise temperatures, expiry dates, and ingredients are missing. This is becoming more pronounced now that the green agenda dominates, requiring precise tracking of the quantity of individual components.

What about order automation? Are Slovenian companies ready for algorithmic demand forecasting that would trigger B2B orders automatically, without human intervention?

Fully automated B2B ordering without human involvement is not science fiction from a technical standpoint. The far more demanding task is algorithmic demand forecasting. That relies on advanced statistical models. We calculate probabilities based on historical data, and the key question is how accurate such a forecast can actually be. It depends on what we want to predict. Satisfying basic human needs is fairly predictable; human whims are considerably less so. Who can say with certainty what someone will want for lunch today?

The technology for order automation is relatively straightforward to implement, but Slovenian companies are largely not ready for it. The reason is simple: companies do not have enough data. They think about automatic stock replenishment, but to get there they first need the right infrastructure in place — you must collect data for processing, ensure its exchange, and establish feedback loops. Only then can automation operate on autopilot. Along the way there are control points where the algorithm can self-correct within defined tolerances. The advantage of the machine is clear: automation will never order something "precisely wrong," whereas a human does so with ease.

Labour shortages are becoming a constant. Do you believe market pressure will force Slovenian companies into warehouse automation faster than they are organisationally and informationally prepared for?

The answer is both yes and no. Understaffing is a fact. You cannot simply import workers in unlimited quantities; you can only have a worker you have developed and trained. A company can do only as much work as its people can perform, or as much as the capital — tools and money — it invests in its processes allows.

Today we face a paradox: a shortage of labour alongside continuously growing demand for services. If kerosene ran out, all unnecessary travel would disappear immediately. The same logic applies to activities that people do not strictly need for survival, yet which continue to thrive. We also face psychological barriers — working in a warehouse or a quarry is considered dirty work. Look at retail: in Slovenia alone, employees manually carry thousands of tonnes of goods onto shelves every day, and customers then take home between 10,000 and 15,000 tonnes of it daily. And then we wonder why there is a staff shortage. It simply is not glamorous to work in a shop or a warehouse, and the pay is poor.

Labour shortages in warehouses are a historical constant — people by nature do not want to do heavy physical work. Other systems solved this in the past through castes; today, in a free society, we must solve it through technology and automation. The market is inevitably pushing companies toward faster warehouse automation, because importing labour does not reduce the problem — quite the opposite, but that is a conversation for another time. The calculation companies make is straightforward: once the cost of an employee reaches a breakeven point — somewhere around €35,000 gross, fully loaded, including every associated cost — the finance team sits down and starts calculating the ROI of automation. And once you automate a process, you eliminate one major staffing challenge for good.

One more thing worth adding: labour shortage is a purely economic problem. Look at the culinary world — everyone wants to be a chef today. There is no shortage of cooks in prestigious restaurants; the shortage is in care homes for the elderly.

Where do you see the limits of warehouse automation in Slovenia? What actually makes economic sense to automate, given the small market size and the specific nature of goods flows?

If you ask strict economists, nothing makes sense at first glance. Economics is a system that takes from one place, gives to another, and keeps the difference for itself. Fortunately, economics is a consequence of human behaviour, not its starting point. But before I digress — to answer the question directly: in Slovenia, the power of example matters. Once someone builds a genuinely world-class automated warehouse here and makes it visible, others will quickly follow and replicate it. Market size is not the critical factor. Our biggest problem right now is that we cannot even solve basic logistics challenges, such as the endless and unnecessary movement of goods. If you reduce the number of handling stages for the same cargo from five to three, you have physically moved three tonnes instead of five. That is genuine rationalisation.

Today, because of poor optimisation, we make an enormous number of unnecessary movements and kilometres searching for specific products across different stores. Personally, I would immediately adopt an advanced shopping application that told me precisely where to buy what, so I drive 30 kilometres instead of 100 — or so that someone else makes that purchase for me. In reality there are no limits to automation. In the long run, everything makes economic sense; in the short run, almost nothing does. The problem is simply that companies are still afraid of large upfront investments.

Any final thoughts on what is essential for understanding the future of logistics?

I am a strong believer in order and standardisation. If you have a material flow established from point A to point B, you need logistics. But logistics can only function effectively in an organised environment, governed by predefined standards — everything else is chaos. From order and standardisation comes mutual trust, which is essential for supply chains to operate. And let us not forget: logistics is a specific discipline that — despite what many believe — is not something everyone simply understands. For logistics, you need instinct. Much like having an ear for music.