In spare parts warehousing, every minute counts. Orders keep coming in right up to cutoff time, while containers still need to be packed and ready for shipment. To reduce peak workloads and time pressure, KION’s in-house logistics provider Urban-Transporte is leveraging AI. A predictive model helps streamline operations and takes pressure off employees on the floor.
2026-06-03
4:00 p.m. in the spare parts warehouse: the last orders of the day are still coming in, while the truck for pickup is already waiting outside. Containers need to be closed as quickly as possible, items picked, packed, and made ready for shipment – under intense time pressure and with absolute precision. It’s the most stressful part of the day for employees, and also the moment when smooth operations and efficient process control matter most.
This is exactly where a new project by KION subsidiary Urban-Transporte comes in. The goal: less stress on the warehouse floor, more stable processes, and greater efficiency in the spare parts business. Using machine learning, Urban aims to predict whether an order is truly complete and ready to be packed – or whether the customer is likely to add more items before the cutoff time.
In the spare parts business, customers often place orders right up until the daily cutoff – the point at which shipments can still be dispatched the same day. Until now, this meant a lot of waiting for warehouse staff. As long as it wasn’t clear whether additional items for the same container might arrive, it couldn’t be sealed.
The result: sharp workload peaks at the end of the day, high operational pressure, and a more complicated handoff to shipping.
“Our customers can place orders until 4:00 p.m., but the truck arrives at 6:00 p.m.,” explains Johannes Neff, project lead at Urban-Transporte. “In those 120 minutes, every box has to be closed, packed, and ready to ship.”
To improve planning and reduce stress, the team developed a data-driven prediction model based on machine learning. It uses Extreme Gradient Boosting (XGBoost), an algorithm that analyzes historical order data and operational variables to generate highly accurate forecasts, including:
• Time of day
• Day of the week
• Public holidays and days leading up to holidays
• Real-time order activity
• Individual ordering behavior
Based on this data, the system calculates the probability that additional items will arrive for a given container. Employees receive a real-time recommendation on whether it can already be closed.
“On a display, the employee immediately sees: no further orders expected – this box can be closed,” Neff explains.
The results from a pilot project at RDC Nordics in Sweden – KION’s own spare parts distribution center for the Scandinavian markets – are impressive: the model achieves an accuracy rate of around 96%.
For warehouse staff, the impact was immediate – and entirely positive:
• Orders can be released for shipping earlier
• Packing work is distributed more evenly throughout the day
• Significantly less stress during peak periods
• Better use of resources
• Smoother overall operations
One particularly clear indicator: the number of packages processed simultaneously dropped from around 70 to about 40. “We were able to significantly level out the entire process and finish earlier at the end of the day,” says Neff.
The project is now also being rolled out in the Czech Republic and Spain.
The biggest benefit is felt by the warehouse teams themselves. Instead of intense pressure right before shipping, packing can start earlier and be spread more evenly throughout the day. This reduces stress during critical windows and improves day-to-day planning.
The physical work remains the same – goods still need to be moved. But travel distances are shorter, errors are reduced, and employees can work more productively. Here, AI isn’t replacing people. It’s supporting better operational decision-making.
The pilot project demonstrates how data analytics and machine learning can make operational processes more efficient and stable. With measurable gains in productivity, process reliability, and resource utilization, the initiative highlights the potential of AI-driven solutions in logistics operations.
Following the successful pilot in Sweden, the project is now also being rolled out in the Czech Republic and Spain.
“With this pilot, we’ve shown how innovation emerges in day-to-day operations – by questioning existing workflows and applying new technologies,” says Neff. It positions the company as a modern, forward-looking logistics partner within the KION Group.
Urban sees further potential for AI in areas such as transport planning, workload forecasting, warehouse slotting optimization, and automated document processing. The goal remains the same: make processes more efficient and noticeably ease the workload for employees.
The project clearly shows: at KION, AI is no longer a future concept – it’s already a practical tool for optimizing real-world warehouse operations today.
At KION subsidiary Urban-Transporte, a pilot project explored the use of AI in spare parts warehousing. The result: the system can predict whether a customer is likely to place additional orders before the cutoff time – significantly reducing pressure on warehouse staff.
Yes. KION is considered a pioneer in the use of artificial intelligence. The company applies AI-powered technologies across vehicles and workflows, including optimizing operations in its own spare parts warehouses.
AI can help prepare e-commerce orders for shipment faster and more efficiently. A KION pilot project at a spare parts warehouse in Sweden demonstrated clear gains in efficiency and a noticeable reduction in workload for employees.
