Warehouse in warehouse model
At a glance:
- Warehouse is a part of the logistics system that stores products (raw materials, work-in-process, finished goods, etc.).
- Warehouse focuses on maximizing the usage of available storage space and controlling the mismatches between system and physical goods.
- When I was a warehouse administrator, I encountered trouble with a small item which had both high stored volume and high order frequency. It was when I came up with an idea about warehouse in warehouse model.
- The essence of this model lies in separating the item into 2 locations, 1 location for daily order and one acting as a fulfillment center.
- The model helped me a lot in controlling and reducing mismatch of this item.
Warehouse is defined as a part of the logistics system that stores products (raw materials, parts, work-in-process, finished goods…) at and between points of origin to points of consumption. While distribution centers’ main purpose is to maximize the throughput by emphasizing the quick moment of the products, warehouse focuses on maximizing usage of available storage space and controlling the mismatches between system and physical goods.
In 2019, I was a warehouse administrator at DB Schenker. We worked on a signed project with Intel called the A9 project to run and manage an internal warehouse locating in their plant for them. There were more than 2000 SKUs in raw material warehouse so it was impossible for us to use static slotting (the practice of assigning a permanent location in a warehouse for each product). Hence, we used random slotting (designating inventory to multiple picking locations) instead.
Everything had been fine till I encountered trouble with a small item which had both high stored volume and high order frequency.
The item was very small and participated in a huge number of transactions every day. In addition, because Intel kept a high stored volume of this item, the random slotting model, which designed this item to multiple picking locations, caused lots of difficulties for us to keep track and manage, hence mismatch often happened with it. Centralizing this item to one location was also considered but it was too hard to recount, check and monitor it. It was when I came up with an idea about the warehouse in warehouse model.
The essence of this model lies in separating the item into 2 locations, 1 location for daily order and one acting as a fulfillment center. First, I collected the total amount of the item and put them into one carton. After checking and making sure that there was no mismatch between system and physical goods, I created a new location for it called “WH” on the system. Then I generated a forecast for the demand of this item in the next day, relocated the forecast amount from the original carton to a smaller one, which would be positioned in THE physical location of WH”. The remaining one was put in the red zone (area for low frequency item) to reduce the transactions of it as lowest as possible.
The WH was placed in a location which bred convenience for retrieving and recounting. Moreover, our system had been set up a rule to generate order to the location which had smallest amounts of item so that every order would drop to “WH”. “WH” was also fulfilled every day using the daily forecast. If there was any shortage, administrators must be alerted and they would make an ad-hoc stock transfer from fulfillment center location in the red zone. The logic behind this model was that I would create 2 areas: one for picking, which had small amount to simplify control, manage and investigate if any mismatch happening; another one would be mainly for fulfillment and storage which had as lowest as possible transactions.
By being monitored closely, the model helped me a lot in controlling and reducing mismatch of this item. It also reduced the effort I had to make to investigate and fix mismatch due to the smaller picking zone.
