Case Studies - The following case studies are adaptations drawn from experience. They are intended to illustrate the application of i-Lean in different practical situations.


i-Lean Case Study - Redesigning a Distribution Center


Client

A parts distributor has a national DC that supplies many smaller DC's with over 20,000 SKU's ranging widely in cost, size, demand, and vendor lead time. Their space and storage media is poorly utilized, material visibility is limited, stockouts are common, material handling efficiency is sub par, and there is no capacity for expanding the SKU base. The company has an old WMS with limited functionality and questionable data quality; material handling activities are often conducted outside of the "system." The company is also interested in supporting a new line of products - which is the impetus for the redesign.


Objectives and Constraints

The intent of this design effort was to:

  • Make better use of existing space and storage media.
  • Improve picking and replenishment efficiency.
  • Reduce inventory carrying and material handling costs.
  • Swap out inefficient storage media.

Project Approach

As a subcontractor to the engineering firm responsible for the DC redesign, Chiron used i-Lean to support the overall redesign effort. This included several steps:
  1. Educating the engineering staff in the capabilities and data requirements of i-Lean.
  2. Working with the engineering firm to gather bulk data from the customer, e.g., SKU information, and derive other reference data, e.g., unit material handling costs.
  3. Constructing an i-Lean database and validating its content for both its accuracy in representing the physical DC environment and its internal data consistency.
  4. Executing basic designs, unconstrained by SKU dimensions and storage media capacity.
  5. Executing advanced designs including SKU dimensions and variable storage capacity.
  6. Evaluating and redesigning picking subzones within the DC to optimize material flow, bin size, and storage utilization.
  7. Documenting design alternatives and associated cost reduction opportunities.
  8. Producing SKU slotting recommendations for the final design.

Data Initialization

As is often the case, this customer had some data readily available in existing systems including SKU volumetrics, demand, and cost. Lead time data was calculated based on available purchase order files made available later in the analysis process. Material handling and other cost factors were derived using a variety of simple techniques. Each data load included automatic validation. After successfully loading all of the data, additional i-Lean validations confirmed integrity across data categories. A small fraction of SKU's with bad or inconsistent data were excluded from the analysis pending correction by the customer. The lead engineering firm further validated the data with the customer to ensure buy in.


Analysis Technique

Once the database was set, dozens of analysis and design iterations of were performed. These evolved from simple to complex. Initial runs addressed the entire SKU population to get an overall view. Subsequently, breakpoint analyses employing demand, lead time, picking profiles, volumetric, and storage media data led to creation of subzones with the desired performance characteristics.

Analysis Set 1 - Match demand for all SKU's with the types of forward pick storage media available based on carton cube assuming a flow path directly from receiving to forward with no reserve storage required. This set of analyses resulted in the best alignment of SKU's with pick storage media, regardless of the amount of media available - a Greenfield approach. Insights included assignment of SKU's to specific media types, e.g., flow rack, shelving, pallet rack, etc. Given the existing facility layout, SKU's were assigned to individual subzones and sorted on Cube Per Order Index (CPOI) as a guide for slotting. This set the stage for more detailed analysis.

Analysis Set 2 - Impose constraints implied by SKU carton volumetrics and available storage media configurations. These analyses fit SKU's into the most appropriate media in a specific subzone based on their sizes and priorities for specific media. For example, 10" high or shorter cartons were assigned to a storage medium to accommodate their height plus shelf clearance of an inch. Once this storage capacity was depleted, remaining Items of lower priority were marked for review. Cartons taller than 10" but shorter than 18" were subjected to the same process. Thousands of SKU's targeted for this subzone were processed in a matter of minutes. In the end many items were assigned to the storage media available requiring only replenishments from receiving; some, however, were not. The process was repeated for the remaining subzones, each of which had certain characteristics, e.g., flow rack for broken case picks, first and second level rack for case picking, floor pile for pallet picks, restricted areas for expensive and hazmat items, etc.

After these SKU assignments were completed, there was still some fallout. Detailed analysis confirms that the remaining items were not assigned to the target subzones for 3 reasons:
  1. The carton was too bulky to fit.

  2. Its priority was lower than those already assigned.

  3. Vendor lead time dictated that direct replenishments to pick slots from receiving could not accommodate demand given the maximum number of slots available for a single SKU (a parameter set for the subzone). The net was that demand exceeded pick space allocated for vendor-based replenishment, indicating that reserve storage would be required.

This fallout led to the next set of analyses.

Analysis Set 3 - Allowing forward picking slots to be replenished directly from receiving (Analysis Set 2) or from reserve storage. This series of analyses addressed problems 2 and 3 from above. Allowing replenishments from reserves had the impact of reducing bin size and ROP for many items since lead time for replenishments was reduced from vendor lead time to a few hours. This freed up space to slot many more SKU's based on optimal flow path.

Some SKU's were still not slotted, however. Typically these were broken case picking SKU's with case dimensions that would underutilize storage capacity, e.g., a 10" high case in flow rack with 18" clearance. Since the customer did not want to re-configure shelf units to generate more 10" shelving capacity, the 18" capacity was made available for smaller items. This technique was applied generally throughout the facility, requiring the movement of some SKU's from one subzone to another. Tools within I-lean made this an easy task. Virtually all of the fallout was resolved.

Analysis Set 4 - Inventory adjustment opportunities. With target bin sizes in hand, a comparison against existing inventory balances, independent of location, was performed. These analyses were based on extractions of data produced from the previous design runs. Some items were under stocked while others were over stocked. The first group could lead to short picks and required immediate attention. The second group, overstock, accounted for hundreds of thousands of dollars in carrying costs alone, not to mention the real estate they consumed. This information was of particular interest to the customer, providing an unexpected boost to the original ROI predictions.


Conclusion

This effort resulted in a plan that yielded a road map for the customer to:

  • Improve picking efficiency.
  • Take advantage of more efficient replenishment strategies.
  • More accurately define bin quantities and ROP's
  • Free up real estate and undesirable storage media
  • Reslot the facility.
  • Identify specific inventory exposures.
  • Provide input to purchasing for improved order management.


In addition, the process of creating an integrated database capturing a sophisticated view of their operation positions them to pursue additional cost reductions and adjust more quickly to changes in their business.



 

 

Top