Supply Chain Segmentation

When studying the incredible success of the Toyota Production System and Dell’s fantastically tight supply chain it is easy to lose sight of the fact that we are looking at very highly evolved operations that have grown organically and dynamically. Dell himself is quoted as saying that in high tech, you either grow or you die. [1]

What are the considerations when instead we wish to optimize an environment where growth is slow or near zero, as might be the case in a stable and well managed hospital or health system? I started initially with questions about how supply chains scale and was led fairly rapidly to studies of supply chain segmentation.

Rather than looking at evolution of an entire supply chain, as one might see in a manufacturing operation with explosive growth, studies of supply chain segmentation turn their attention to where an operation would be well served to break up its supply chain for various groups of items on the basis of certain item properties such as price or size.

In his paper Differentiating the Hospital Supply Chain For Enhanced Performance [2] from the Engineering Systems Division at MIT, DeScioli submits that hospitals need to have more than one supply chain and performs simulations that support his thesis. On the basis of his results, he proposes that hospital supply chains be segmented on the following bases:

• Unit Price: Expensive inventory items requiring tight control should be kept in locked cabinets accessible only by authorized login, whereas items not requiring should be kept on open shelf systems to minimize staff overhead in retrieval of the item in question. 
• Unit Size: Large items have a larger carrying cost than smaller items of similar cost simply because they occupy more physical plant. As a result, large items require near continuous review of need for inventory on hand whereas such review is less critical for smaller items with correspondingly lower carrying costs on the ward.
• Unit Criticality: The cost of stock-outs for critical items should be evaluated in order to plan appropriately for inventory needed. DeScioli proposes a three-tiered system of measurement of criticality and corresponding storage, e.g. open storage for non-critical items and password-controlled locked storage for highly critical items.

In a later paper from the same department, Cheng and Whittemore [3] take best practices gleaned from studies of supply chain segmentation in other industries and apply them to the hospital setting. They propose a “Next Generation Hospital Supply Chain” where inventory items move directly from a staging warehouse to the bedside of a specific patient on the basis of that patient’s diagnosis and clinical status. These “personalized kits” reduce overhead minimizing space for inventory on the ward and by reducing time required by clinical staff to retrieve needed items.

Question:

What is the break-point where it makes sense to consider supply chain segmentation for any given entity? For small entities, formal implementation of segmentation policy simply adds confusion and overhead. For large entities there are enormous benefits to carefully considered segmentation. How does an entity know that it is time to begin to implement segmentation policy?

References: 

[1] Breen, B., (November 1 2004). “Living in Dell Time.” Fast Company http://www.fastcompany.com/51967/living-dell-time accessed 17 September, 2012.

[2] DeScioli, D.T. (2005). Differentiating the Hospital Supply Chain For Enhanced Performance. (Masters Thesis). Retrieved from http://dspace.mit.edu/handle/1721.1/33317

[3] Cheng, S.H.; Whittemore, G.J. An engineering approach to improving hospital supply chains. (Masters Thesis). Retrieved from http://dspace.mit.edu/handle/1721.1/44928