Introducing CAPcargo AI LoadPlanner

AI should not be about being fancy. It should deliver real business value.

At CAPcargo, we were not satisfied with the results of the static load planning approach we had been using. So, we challenged ourselves: Could AI make a real difference? Kann KI tatsächlich einen Unterschied machen?

The answer has been surprisingly positive.

Our new CAPcargo AI LoadPlanner* is delivering better loading results within an acceptable processing time. And there is another aspect that makes this particularly interesting: developing a comparable module using traditional software development methods would probably have taken us 12–18 months or more. CAPcargo AI LoadPlanner* liefert bessere Beladungsergebnisse innerhalb einer akzeptablen Verarbeitungszeit. Besonders interessant ist dabei ein weiterer Aspekt: Die Entwicklung eines vergleichbaren Moduls mit herkömmlichen Softwareentwicklungsmethoden hätte vermutlich 12–18 Monate oder länger gedauert.

But this is only the beginning.

We are continuing to explore how the system can learn from the corrections made by human planners, creating a more adaptive and continuously improving solution.

Another major advantage of the AI-based approach is flexibility. We can adjust and extend the underlying logic much more easily to accommodate specific use cases — for example, defining when package overlaps are acceptable or when a package may be tilted.

For us, this is a great example of what AI should be about: not technology for technology’s sake, but a practical tool that solves real operational challenges and creates measurable value.

And yes, we are excited about where this journey will take us. 🚀

AI is not replacing the expertise of our transport planners. It is helping us turn that expertise into a smarter, more flexible planning engine.

* A load planner application calculates how cargo can best be arranged inside a transport vehicle. Internally, we call it the “Tetris problem.” „Tetris-Problem“.