BY HANS BERGGREN, AGNETA CHRISTERSDOTTER, AND JENS DREMO
Download the English version here
The supply chain has always been about interaction: between demand and production; between inventory and delivery; and between companies and operations that, in practice, depend on each other’s decisions but are often guided by different incentives, systems, and time horizons.
That’s also why the supply chain is changing a bit more slowly than many other sectors and industries. Not because the technology isn’t there, but because the reality is complex. There are many players with long-standing and complex relationships. Add to that legal structures and business-critical dependencies, and you get a world where day-to-day stability is often valued more highly than experimentation. At the same time, something is happening right now in our world of supply chains. Not in the form of a dramatic revolution, but as a cautious curiosity about how technology can be used in entirely new ways. As the level of digitized flows in supply chains really begins to take off, we can start using predictive AI as a way to better understand the pace at which the market is already moving.
Forecasts aren't the problem—uncertainty is the problem
Forecasting is a part of nearly every supply chain. Often done manually, sometimes supported by systems, and almost always with significant margins of error. This is particularly true in industries with long lead times and complex interdependencies, such as the automotive industry, where forecasting has been used for over 30 years and the volume of shared data is very large. The problem lies in the consequences of the uncertainty surrounding these forecasts, which manifest as excessively high safety stock, capital tied up unnecessarily, late changes that lead to last-minute scrambles in production, or, in the worst case, scrapping and lost delivery capacity.
This is where predictive AI really starts to get interesting. Not because it eliminates uncertainty in the supply chain—that will always be there—but because, in certain areas, it can reduce that uncertainty. This is especially true when, as in the automotive industry, the volume of data is vast. When the range of uncertainty in the decision-making process shrinks—even marginally—it changes how people plan, order, and produce. For certain items and flows, that improvement alone may be enough to create a calmer, more synchronized behavior throughout the entire chain. We have seen this clearly in our work with the automotive industry’s supply chain. When forecast signals are analyzed over time, in combination with historical patterns, a more stable picture of demand emerges than the one often communicated in traditional forecast files. The effect isn’t dramatic at every data point, but it’s sufficient to influence inventory strategies, purchasing decisions, and production planning.
Synchronization is at least as important as optimization
For many years, supply chain development has been about optimization: lower costs, shorter lead times, and higher service levels. These factors remain important, but as complexity increases, synchronization becomes at least as crucial. A synchronized supply chain isn’t about everyone doing the same thing, but about everyone moving in step. It’s about decisions being made based on a shared understanding of reality. It’s about changes not coming as surprises, but as signals that are picked up in time. This is where the orchestra metaphor comes in handy. In an orchestra, not everyone plays the same instrument, but everyone follows the same beat. The conductor sets the tempo—and in the supply chain, it is effectively the market, in combination with key stakeholders who can influence the tempo, that acts as the conductor. Demand, economic conditions, customer behavior, and external disruptions set the rhythm. The question is how well organizations and networks manage to keep up with it. In this context, predictive AI can be seen as a way to hear the beat more clearly, rather than trying to rewrite the sheet music.
Trust, Data, and Human Decisions
A recurring issue in the use of AI in the supply chain is trust—not in the technology itself, but in the decisions it leads to. Many organizations are more forgiving of human errors than of technical ones. An incorrect forecast made by a planner can often be explained and accepted. An AI-based recommendation that deviates from the norm, on the other hand, raises concerns, even if it is statistically superior. That’s why the adoption of AI is at least as much a cultural journey as it is a technical one. Transparency in how the models reason, the ability for human oversight, and clear ”learning loops” are crucial. In practice, predictive AI works best when it doesn’t replace decisions but challenges them—and when the organization allows itself to learn from the outcome. This has been a central principle in PipeChain’s work: AI-based forecasts are sound recommendations, not a magic crystal ball. It is still people who take responsibility for the decision, but with a broader and more consistent foundation. We believe in automation with control and regulation functions that ensure the best possible results and value creation over time.
Small Steps, Real Problems
Another important lesson is that value creation rarely results from large, overarching AI initiatives. Value is created when the technology is applied to concrete, everyday problems and a cultural shift begins. Our own AI initiatives are often employee-driven and focused on solving specific challenges our customers face. This might involve interpreting incoming PDF orders that would otherwise require manual processing; automating the compilation of customs documents; or gradually improving forecast accuracy to reduce the risk of overstock or stockouts. The common thread is that technology isn’t introduced as something abstract, but as a way to reduce friction in existing workflows. That’s also where acceptance grows: when people realize that their work actually becomes easier, calmer, and more predictable.
From Hype to Everyday Life
AI in the supply chain is neither a passing trend nor a one-size-fits-all solution. The real change lies somewhere in between. We believe that the organizations that will succeed in the future are those that combine realism with curiosity: those that dare to use technology where it’s useful, but don’t expect miracles. Those that invest in data quality, relationships, and ways of working—not just in models.
Ultimately, it’s about building supply chains that can better handle variability. Chains that react faster, but not more erratically. And that, as the market changes pace, manage to keep up without losing their balance. That’s where a synchronized supply chain begins to take shape. Not perfect. But in step.
—
Hans Berggren is the CEO of PipeChain
Agneta Christersdotter is the CEO of PipeChain Networks
Jens Dremo is the CEO of PipeChain SCM Tyringe
