Introduction
Designing a perfect warehouse is a concept that exists only in books. It is almost impossible to find a perfect solution and then implement that with real life constraints. A logistics network has so many parameters and variability that it is not possible to design a perfect warehouse (Harrington, 2007). Designing a logistics network with many suppliers, customers, warehouses and plants is exhaustively detail-laden. Among all types of logistics network, design with one central network design is mostly used. With one warehouse logistics network, many theoretical aspects of network optimization are often easy. This paper will discuss the design of a warehouse in a logistics network.
Objective
Designing any logistics network is solely dependent on the objective of such exercise. This also depends on the business objective of the company. The major objectives that usually drive the network design include the following:
Service Level of Delivery to the Customers.
Total Cost of Transportation.
Facility Cost ( includes handling cost and fixed cost of operation)
Inventory Carrying Cost
Initial Cost of investment to build the Warehouse.
One among the above five or more than one of the five drive the design decisions (Harrington, 2007).
The main objective of selecting a warehouse is to pick a location that is close to most of the customers who will improve the service level of delivery (Harrington, 2007). Service level in most of the cases is measured in terms of lead time to the customer from the warehouse. It is important to first understand the objective fully by trying to find a location. For example, the objective of a company may be to achieve a service level of 3 day delivery to the customers in 90% cases.
Once the objective is finalized, it is important to identify the nodes of the network and categorize them accordingly (Chopra, 2011). Mapping all customers or suppliers may be ideal but if they are too many in number, then aggregating customers and suppliers is a better idea. Aggregation according to geographical proximity is the most common method used.
Aggregation of the products may also be required as network optimization with too many products may be cumbersome and data collection may not be feasible (Chopra, 2011). Products based on similar characteristics are generally aggregated. However, while doing product aggregation, product priority is also taken into account to give more weightage to high priority customers.
Once nodes are identified and products are determined, data collection for the network begins. Accurate and exhaustive data collection is key to a successful network optimization and coming up with a solution. Parameters like inbound and outbound transportation cost, inventory level, facility cost are some of cost that are required to run network optimization to determine a location (Chopra, 2011). However, apart from these real life costs, some notational costs are also important to give input regarding how important the fulfillment of the demand is compared to the non-fulfillment of the demand. Delay cost, non-delivery cost are something that network designers should determine based on the objective of the company (Chopra, 2011).
Once the nodes, products and parameters are given as input, network optimization can be done. In some cases, some companies determine the costs for a number of potential warehouse location options and then select the one with the lowest cost and highest service level. In other cases, optimization is done to determine which geographical location is best suited to build a new warehouse. There are several statistical techniques available for optimization such as ant colony optimization, genetic algorithm, geo-coordinate optimization, linear programming and P-median algorithm (Vidya and Maheswari, 2011).
Warehouse Design
Once the location is finalized, it is important to design the warehouse itself. The first step of designing a warehouse is to collect operational data. The three major areas of a warehouse design are size and capacity, shipment handling capability, and storage bay design (Chopra, 2011).
Operational data vital to warehouse design are what will be the inbound versus the outbound flow of materials, what level of inventory should the warehouse maintain to provide a desired level of customer service, (Safety Stock Level) and how much of the inbound delivery can be retransferred to another shipment without the need for storage versus what percentage needs to be palletized and stored (Chopra, 2011). Before optimization, it is important to gather cost data for handling, labor cost, storage cost, inventory carrying cost and fixed cost of running the warehouse. Apart from cost parameters, other operational parameters such as lot size and rounding value are also required for warehouse design optimization. Inbound delivery should be given more importance in warehouse design decision as inbound delivery is dependent on other logistics partners and often have a higher level of uncertainty than outbound delivery, which is within the control of the warehouse (Vidya and Maheswari, 2011).
Once data is gathered, warehouse optimization statistical models can be run to determine the size of the warehouse (Vidya and Maheswari, 2011). However, unlike single location logistics network, designing a single warehouse for a network that has no other storage point is very complex. There are also other parameters that often cannot be modeled such as green pallets, mobile equipment, sorting equipment, picking modules, staffing and capital budgets. Also, there are other things to take into consideration. If the design of the warehouse is flexible, how well suited the warehouse is for both mechanical and automated operation and how difficult for the workers to understand and work in the warehouse are some of the essential considerations (Vidya and Maheswari, 2011).
Conclusion
One warehouse logistics network may seem an easy task to design, but in real life, it is not a cinch. In the case of a single warehouse network, it is even more important to design in the perfect way as the logistics network will have no alternate options in case of a failure. Determining a location that will optimize the service level and network cost is essential. The design of the warehouse should be such that it optimizes the inventory level and operational cost.
References
Harrington, L. (2007). Designing the Perfect Warehouse. Inbound Logistics. Retrieved on 19 April, 2015 from <http://www.inboundlogistics.com/cms/article/designing-the-perfect-warehouse/>
Graves, S., Bhatnagar, R., & Cheong, M. (2007). Logistics network design with supplier consolidation hubs and multiple shipment options. JIMO. Retrieved on 19 April, 2015 from <http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=3&cad=rja&uact=8&ved=0CDEQFjAC&url=http%3A%2F%2Fweb.mit.edu%2Fsgraves%2Fwww%2Fpapers%2FJIMO%2520paper%2520Michelle%2520Rohit.pdf&ei=YDI0Vbf1OIOXgwSo8oDIAQ&usg=AFQjCNGV3b-LLpyHl3Ayq0RHNbOvPwFlSg&sig2=IsuUcZKpPEU1PXeBFa_UFA&bvm=bv.91071109,d.eXY>
Chopra, S. (2011). Designing the Supply Chain in a Distributed Network. Kellogg’s Business School. Retrieved on 19 April, 2015 from < http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=10&cad=rja&uact=8&ved=0CGcQFjAJ&url=http%3A%2F%2Fwww.transportation.northwestern.edu%2Fdocs%2Fresearch%2FChopra_DistributionSupplyChain.pdf&ei=YDI0Vbf1OIOXgwSo8oDIAQ&usg=AFQjCNFweR-UgSiqepAe7548xKothbzC3g&sig2=QntKyyvpFd795bKDcxMSSw&bvm=bv.91071109,d.eXY>
Vidya, P. and Maheswari, L. (2011). Designing the Distribution Network in a Supply Chain. IJAR, 4(5), 367-369. doi:10.15373/2249555x/may2014/110.
The Optimization Research of Supply Chain Structure Based on Supply Chain Competition. (2013). JCIT, 8(5), 1220-1228. doi:10.4156/jcit.vol8.issue5.141
