Bleckmann, a contract logistics specialist for fashion and lifestyle brands, said on November 17 that it is piloting Carrousel AI, an AI-powered robotic picking arm, at its site in Bergen op Zoom, the Netherlands. In a company announcement, Bleckmann said the three-year pilot is a collaboration with AutoStore and Berkshire Grey, and that the arm has been used in the AutoStore picking process since it was installed in mid-June. After months of testing and verification, the company said, the system is now “running smoothly and reliably.”
Bleckmann said the arm uses real-time sensor and visual data to carry out picks, record results, and adapt its methods, including to new stock-keeping units and to changes in stock. With suction and vacuum tools in various sizes, the company said, the system can handle a wide range of products, including awkward or bulky items, and can change suction cups on the fly based on what it has learned about each product. In the first few days after the arm went live, Bleckmann said, it picked up to 350 units per hour, a rate the company said exceeded its expectations for the start-up period.
“At Bleckmann, we embrace innovation and invest in it to provide future-proof supply chain solutions. And, of course, we innovate to keep pace with our customers’ changing needs. Carrousel AI is again a key milestone in our automation and growth strategy,” said Kevin Paindeville, director of warehouse solutions and innovation at Bleckmann. The company said automation is a pillar of its strategy for fashion and lifestyle customers, and that automating labor-intensive picking can help those customers benefit from later cut-off times for next-day delivery. It also said the arm, like AutoStore, is part of an integrated setup that can operate 24 hours a day.
Those operating claims, and the 350-units-per-hour figure, come from Bleckmann. The announcement does not report an independent throughput study, a sustained rate after the first days, an error rate, or a comparison with manual picking at the same site.
What it means for manufacturers
A midsize manufacturer that does not run a fashion fulfillment center may still recognize the problem: mixed SKUs, odd shapes, and a picking process that is hard to staff around the clock. A three-year pilot is a reminder that even a specialist 3PL is treating adaptive picking as a test, not a finished rollout. The more useful questions are whether the arm can keep up after the novelty period, and whether exceptions still have a clear path back to inventory records — the same issues discussed in warehouse automation for midsize plants.
A practical next step is to time a representative pick wave, note the share of lines that need a second look, and write down what happens to the inventory record when a pick fails. Recent coverage of DHL’s adaptive robotics work and GXO’s Apollo trials offers two other operator-led examples to compare before anyone issues a purchase order for a new arm.
Questions about pick-wave performance, warehouse exceptions, or whether a robotics pilot is worth scoping?
Contact usSources
Additional sources
- As of this writing, the Bleckmann announcement is the primary public account of the Carrousel AI pilot, including the mid-June installation, the AutoStore and Berkshire Grey collaboration, the three-year term, and the first-days figure of up to 350 units per hour. No independent throughput study of the Bergen op Zoom installation was located for this report.
Photo courtesy of Bleckmann, from the company’s Nov. 17, 2026 announcement.
Try the Statadventure live tools
Inventory tools for US manufacturers: reorder point, safety stock, EOQ and landed cost, with the inputs and assumptions shown.
Open the live tools