Packing planning shipping carton
Initial situation
At the logistics center of Adolf Würth GmbH in Künzelsau, a wide variety of goods are repacked from single-variety storage boxes into differently mixed shipping cartons and then sent to the customer. In order to achieve the most compact packing patterns possible, the packing patterns for the shipping cartons are created in accordance with a series of guidelines.
Repacking goods from a source container into the shipping carton is a very monotonous and tiring task for which it is becoming increasingly difficult to find and retain qualified personnel. New skilled workers, or skilled workers who tire towards the end of their shift, sometimes have problems creating compact packing patterns in accordance with the guidelines and should therefore be supported in finding suitable packing patterns. In the medium term, data should also be collected in order to be able to automate the task in the future.
Solution idea
The packing patterns generated by the staff will be used to train an AI, which will then make suggestions for good storage points. These suggestions can then be displayed on a screen or with the help of augmented reality during the process, helping to reduce the workload for the specialist staff. In addition, the packing planner trained on the collected data can also be used together with robots in the medium term and thus completely automate the task.
Benefit
The packing planner can relieve and support employees in logistics. New staff in particular can benefit from the planner's support.
The data set that would be created by recording the packing patterns laid by humans would be extremely valuable for training an AI planner. Furthermore, the recording of packing patterns could be used for quality assurance, as it can be used to traceably ensure that the right goods end up in the right shipping boxes and that no incorrect or incomplete deliveries are sent to the customer.
Another important point is that a reliable planner can not only show employees good packing positions, but can also be used for automation with the help of robots. This is particularly important in times of a shortage of skilled workers.
Implementation of the AI application
In the course of the Quick Check, the model-free packing planning developed by Fraunhofer IPA was analyzed for the use case at Adolf Würth GmbH. Investigations were carried out into how the packing planning can be integrated into the logistics system on site and what a possible implementation of the tracking and recommendation of the packing pattern could look like.
However, the challenge of connecting to the productive logistics system on site and the necessary disclosure of system interfaces by the system integrator could not yet be implemented with the planned resources, so that the application could not yet be realized within the scope of this project.
During the course of the project, however, another possible application for the pack planner developed by Fraunhofer IPA was identified. This possible application is currently being examined and the joint collaboration is now focusing on this new possibility.