For the centralization of its logistics at its main location in Dachau, the technical trader Ludwig Meister is relying on the implementation of Sereact's AI-driven Pick and Place system. This is stated in a press release from the AI startup.
With around 2.5 million items, Ludwig Meister has a comprehensive product portfolio. The challenge: conventional Pick and Place automation solutions require extensive data that they must be trained with before they can be reliably used.
“We didn’t want to train two and a half million products but rather use the system proactively right away,” explains Matthias Dambach, Head of Central Logistics at Ludwig Meister.
Another major challenge was the gentle and precise handling of the goods. The products include sensitive rolling and precision bearings, which could be damaged by rough picking and thus rendered unusable for installation. Through precise placement of the goods, the system should also be able to efficiently utilize the space of the eighth containers.
The search for
a suitable solution provider finally led to Sereact. At LogiMAT 2023, contact was first made with the Stuttgart software developer after Ludwig Meister visited their booth. A deciding factor for the collaboration was particularly Sereact's expertise in AI-based robotics.
“Other companies we were in contact with could not promise us such a high degree of automation as Sereact,” summarizes Dambach.
Due to the large and diverse product portfolio, the company was keen to first determine which products are suitable for the use of the Sereact solution. In the first step of the collaboration, Ludwig Meister therefore sent part of its items to Stuttgart to assess Sereact’s range of services based on practical gripping tests. On site, the Bavarian service provider also had the opportunity to view the application up close. Only three months later, a complete, ready-to-use Pick and Place system, including a robot cell, was realized at the Dachau headquarters, according to Sereact.
According to the provider, the system enables
space-saving positioning of products and offers the possibility to integrate individual changes of items into the applications afterward, without requiring training. The object-centered Vision Language Action Model (VLAM) allows robots to independently analyze unknown situations without prior training and to carry out appropriate actions.
Natural Language in Operation
Operation is carried out via natural language, allowing robots to be programmed and their behavior modified without any prior knowledge. For example, Ludwig Meister defined an application scenario where bags that are 20 percent larger than the compartment in which they are to be placed can still be stored there, as the packaging is flexible. In another case study, the system was supposed to recognize boxes marked with an “X,” as these have already been opened for individual removal. This information was subsequently instructed into the AI with the input prompt “Do not pick items with a painted-on X,” the report says.
The handleability of some products tested by Sereact was documented and
categorized.
“However, we still had a large stock of untested products, whose handleability cannot be conclusively assessed yet,” explains Dambach. “Therefore, these products are left to the robot cell, which independently decides whether reallocation is possible. The success rate is between 96 and 98 percent – and that with our large portfolio.”
In addition to the AI-based software, Sereact’s solution includes a Triple Suction Gripper specially developed for Ludwig Meister. This is a gripper with three different suction cups. By retracting the unused suction cups while simultaneously picking up the product with the active suction cup, the gripper arm can reach the bottom of the deep containers. For the future, the company plans the integration of another robot cell, which will function as a piece-picking robot and commission customer orders.
“Sereact's solution has helped us better understand our own product portfolio,” sums up Dambach. “With this knowledge, we are able to seamlessly automate the logistics of our headquarters and focus on
