The picking methods employed in warehouses vary depending on warehouse size. Small storage facilities will often pick orders individually while large multi-zone facilities will gather orders in batches. In between, there is a gray area, where orders are complex and warehouses neither large nor small. In these cases, what are the optimal order picking methods?
DHL Supply Chain conducted research and developed an optimization tool for order picking in medium-size warehouses. Here, learn about their research, solution, and case study results.
DHL Supply Chain is a division of Deutsche Post DHL Group with a global network and extensive logistics portfolio, including warehousing, transport, and value-added services.
Problem
Cluster picking is a method for collecting stock keeping items (SKU) for multiple orders in a single assignment. It is commonly implemented with a first come first served (FCFS) wave strategy.
The cluster pick method is used in medium size warehouses instead of picking orders individually, as practiced in small warehouses, or batch picking, which is used in large warehouses. Cluster pick is an attempt at maintaining a good level of throughput while dealing with many orders. At maximum capacity, congestion and inefficiency can become challenges.
Read more about warehouse throughput in the DHL Supply Chain case study.
The Operations Science Team at DHL Supply Chain wanted to improve on basic cluster picking by reducing congestion and improving efficiency. Their solution would be a software tool that could be deployed at medium size warehouses.
Solution
Engineers developed a simulation model of cluster picking in a warehouse. The model was used to replicate the cluster picking method and compare it against alternatives. In this way, the engineers could resolve congestion and other bottlenecks in the order picking processes.
The model accounted for warehouse shifts, the number of pickers in each shift, and gave the possibility to compare alternative waving strategies against FCFS. Metric collection included aisle and cart congestion, cart journey completion time, and waiting times.
Testing showed that the optimal cluster picking solution should focus on reducing picking cart travel distance. DHL’s engineers found that rather than have a cart visit many or all aisles in a warehouse on a single assignment, the optimal method was to arrange orders for collection in a way that would mean visiting as few aisles as possible.
The DHL team further developed their order grouping algorithm beyond minimizing travel distance to include minimization of stops and to balance work across zones when applicable. The result increased the number of units processed per hour, minimized order cycle time, and reduced congestion.
To help apply cluster picking analysis and order grouping methods across the company, a micro-service plug-in was developed for DHL’s warehouse management systems. The system was called IDEA (Instantly Discover Efficient Activities).
Results
DHL Supply Chain’s Operations Science Team validated IDEA, showing an overall increase in productivity of 14% and a decrease in cart congestion of 35% when compared to FCFS. As a result of these efficiency findings, it would be possible to reduce the number of pickers by 12.5%.
The efficiency of IDEA, when compared to FCFS, resulted from improvements in picking time and cart idle time. In the test scenarios there was a 12% reduction in the time required for a cart to complete a picking assignment (cart completion time) and a 36 % reduction in the time carts were waiting for slots to become free.
The comparison of cart congestion for IDEA and FCFS illustrates the reduction in time carts were waiting for slots to become free. Noticeably, IDEA reduces the time more than four carts are congested from 28% of the time to 18% of the time.
Overall, the IDEA tool is an effective way to reduce operating costs by reducing the number of pickers needed to maintain warehouse throughput. The tool is easily integrated with DHL’s warehouse management system as a plug-in and can be deployed wherever needed. It is an example of how simulation modeling with AnyLogic can deliver powerful results in a way that integrates with existing systems. Learn more about warehouse optimization using AnyLogic.
This case study is from a presentation given by Vijay Sharma, of DHL Supply Chain, at the AnyLogic Conference 2021. His presentation with follow-up question and answer session:
