Why are transparency, integrity and reliability so critical in logistics technology? Peter McLeod speaks to experts.
If there’s one theme that most clearly cuts through the noise at this year’s LogiMAT, it’s speed. It’s not just about speed of operations, it’s about speed of deployment, speed of innovation, and ultimately speed of return on investment. For Inform Software, this discussion leads to a broader issue. How can logistics organizations implement more intelligent systems without losing transparency, control, and trust?
Speaking to me at a crowded exhibition hall in Stuttgart, Dr. Bernd Heinrichs, Inform’s senior vice president of inventory and supply chain, outlined how the company sees the development of artificial intelligence in supply chain and intralogistics environments.
Extending the optimization layer
Inform has long been associated with optimizing complex data-driven environments. However, as markets become more volatile, optimization systems are required to react faster, incorporate more signals, and support more dynamic decision-making.
This change is particularly relevant in environments where decisions are interdependent. Changes in demand planning can affect inventory, transportation capacity, labor allocation, or service levels. Transparency is essential for day-to-day use, as recommendations made in one part of the operation can have an impact elsewhere.
For the Heinrich family, this is an area where the practical value of AI in logistics must be proven. “I’m not talking about AI. I’m talking about explainable AI,” he says. “Everything we do and propose has an explanation, otherwise people won’t trust it.”
Trust as a practical requirement
In conversations with customers in various industries, he says the same question comes up over and over again: “Why did the system choose that option and not another?”
This question is important because logistics decisions are rarely driven solely by technology. These involve planners, managers, operations teams, and often customers and external partners. If these stakeholders cannot follow the reasoning behind an AI-assisted recommendation, they are less likely to act on it.
For Heinrichs, this could become a meaningful point of differentiation for European technology providers. “We can build good AI just like anyone else, but we can also add something different,” he says. “It shouldn’t be a black box.”
That distinction will become increasingly important as companies seek to incorporate AI applications into established business processes. The system needs to be technically powerful, but also understandable enough for users to challenge, test, and improve over time.
Managing unpredictable environments
It is becoming increasingly difficult to plan your operational environment based solely on historical data. Demand patterns change, external factors intervene, and market conditions can change rapidly, often before the changes are clearly visible in the numbers. “Instead of relying solely on historical data, we need to collect real-time data,” he says. “You need to react to volatility and incorporate signals from a variety of sources into your decision-making.”
This marks a shift from more static optimization models to more responsive systems that continually consider new information. “It’s becoming more dynamic,” he added. “The next step is to make it more agentic, reacting on its own to changes in its environment.”
From news to predictions
One example of Inform, first announced at LogiMAT, is a new AI-based approach designed to incorporate external events directly into forecasting and scenario planning. Heinrichs said the starting point was a simple question: “Why do predictive models so often ignore what’s happening in the world around them?”
“When you do classic forecasting today, it’s based on historical numbers,” he explains. “But in reality, demand is constantly being influenced by events such as geopolitical conflicts, supply chain disruptions, new regulations and market trends. This information exists, but usually in the form of news rather than numbers.”
The new solution is designed to fill that gap. Users provide time series such as sales figures and market indicators and briefly explain their background. The AI then examines relevant news events, analyzes past relationships, and generates several possible future scenarios. The result is predictions with evidence-based explanations as to why the market is likely to develop in different directions.
humans in the loop
For Heinrich (pictured below), the discussion around AI also connects directly to the role of human expertise. AI can identify patterns, process large amounts of information, and create scenarios quickly. But its value increases when people can add experience, context, and judgment that data alone cannot provide.
“AI is only as good as the data it works with and the people who can give meaning to that data,” he says. “That’s why humans remain an important part of the loop.”
In practice, this means that planners and decision makers are not excluded from the process. They are still the central figures. Their role is to validate scenarios, question assumptions, and refine outputs based on operational knowledge and market intuition.
“Once you understand why a system recommends something, you can decide whether to trust it, question it, or improve it,” Heinrichs explains. “That’s where the collaboration between human judgment and machine intelligence becomes very powerful.”
Integration and interoperability
Another consistent theme in our discussions with customers is integration. As logistics operations become increasingly interconnected, the ability to link AI-driven applications with existing systems is becoming essential. “We get asked all the time: How do we integrate with ERP systems and other solutions?” Heinrichs tells me. Inform’s response was to standardize connectors and integrate with major platforms such as SAP and Microsoft. The result is an easier integration path, reducing both cost and implementation time.
“It makes a big difference,” he added. “And it also makes it easier for us to expand internationally.”
This is an important point in introducing AI. Even the most sophisticated applications have a difficult time creating value when business processes are located away from the systems in which they are actually managed. Logistics companies already operate in established IT environments, and new solutions must fit into those environments without adding complexity.
Data Responsibility
As connectivity and data usage increases, security scrutiny increases. Heinrichs’ background in cybersecurity speaks to his strong stance on this issue. “Every product needs a security stamp before it’s released,” he says. “It’s an obligation.”
As AI models draw on a wider range of data sources, including external feeds such as news and market information, the complexity of managing and securing that data increases. “The amount of data we are working with has created a huge demand from a data security perspective,” Heinrichs points out. “You have to stay on top of that.”
the market is ready to move
Perhaps most impressive is Heinrichs’ assessment of market sentiment. Rather than caution, he sees a growing desire for experimentation and rapid progress.
“Customers come to us for ideas,” he says. “They are willing to win fast and fail fast.” This openness creates fertile ground for intelligent solutions that can deliver measurable improvements without the inertia of large-scale transformation projects.
For many companies, the next phase of digitalization will not be defined solely by AI. It is defined by AI that explains itself, connects cleanly with existing systems, and supports decision-making that people can trust.
