SAP EWM add-ons (1.): Automatic task assignment to the warehouse worker’s reader

Fully automatic task assignment to the mobile terminal across queues using mib:EWM logic.

The first and essentially crucial element of our series on SAP EWM add-ons is the fully automatic task assignment to the mobile terminal across queues using mib:EWM logic. It is absolutely essential because it applies to all warehouse processes. It significantly reduces the need for decision making by the individual warehouse worker and thus prevents errors arising from these decisions. It is therefore a powerful tool for optimising warehouse operations.

We developed this solution back in 2009 as an optimization algorithm within our superstructure over LE-WM, called mib:WMS. Since 2019, we also offer it in the EWM environment and it helps us to solve complex situations in large warehouses. In total, around 25 warehouses across Europe are already using it successfully.

Business case

Warehouse workers should ideally work only with a mobile terminal – a reader. They should log in to the scanner and specify the handling equipment they are using. From there, suitable tasks (warehouse orders) should be automatically assigned to them one by one. It’s essential to dynamically consider all relevant criteria to assign the worker the optimal warehouse order for processing.

The EWM standard offers prioritization within only one selected queue, basically according to the planned completion date of the warehouse order. The warehouse worker must therefore select a specific queue manually and process the orders in a fixed order.

The standard also offers the possibility of switching between queues by a somewhat limited logic called interleaving, where switching to another queue follows a fixed sequence (e.g. Picking – Storage – Picking – Storage). However, this is hardly practical in real operations.

Yes, processing all orders in the queue one by one will eventually reach the goal, assuming everything goes as planned. However, it is possible that:

• someone might finish their queue early and start looking for more work, while another might be struggling to keep up with their tasks
• someone might get delayed in their queue due to an exception (e.g., a picker cannot find the required quantity), causing a delay in subsequent operations.
• it will not be possible to complete the picking and packing of a dispatch order due to a missing item
• priority will not always be given to tasks whose completion is critical to subsequent activities
• the warehouse worker will spend extended time traveling with an empty load
• bottlenecks will occur (e.g. overcrowded area at packing counters)
• completing the entire workload will be delayed due to the issues described above
• some warehouse operations will be carried out with more expensive handling equipment or higher-skilled resources necessary

mib:solution

We believe that only a system algorithm can determine the optimal order for processing tasks ensuring that all resources are properly utilized, all exceptions are captured, and work is completed on time within the required deadlines.

That’s why our solution aims to:

1. not require the worker to manually select a specific queue
2. compile a complete list of warehouse orders for the given user with the specified type of equipment across all queues
3. take into account the characteristics of the handling equipment used
4. prioritize within this list using a set of complex criteria:

• Deadline based on the latest start time for order processing: prioritize tasks that are most urgent and will take the longest to complete.
• Travel time to the starting location of the order: process the order closest to me
• Priority of the entire queue: operationally editable by the manager in the warehouse monitor
• Warehouse order priority: operationally editable by the manager in the warehouse monitor
• Order completeness priority: ensures smooth flow of subsequent operations (e.g. we don’t wait long for the last item to be packed)
• Number of waiting pickers: speed up the replenishment of picking locations to enable uninterrupted picking
• Filling the capacity of the source and destination area: prevent area overfill by limiting the number of handling units or workers
• + any additional customer-specific criteria (e.g., statistical error rate of a picker for certain items, etc.)

Each criterion can be assigned a different weight within the queue. There are queues (e.g. picking) where the main criterion is the deadline. For other queues (e.g. storage and replenishment), we try to minimize unnecessary transfers.

Prioritization occurs before each assignment of a new task to a worker, making it continuous and dynamic. This allows us to respond flexibly to the actual duration of specific tasks compared to the plan, processing exceptions, priority changes, and the real-time location of the worker in the warehouse.

For this purpose, we have also optimized the standard distance calculation (EWM TDC Travel Distance Calculation), which must be performed for a large number of orders simultaneously. However, a description of this modification deserves a separate article, to be shared in the future.

Consultation

Are you interested in a consultation regarding this solution? Feel free to contact us.