- Safety stock and management
- Safety stock and material requirements planning (MRP)
- Safety stock and performance measure
- Safety stock and lot-sizing
- Safety stock and lead-time
In fact, safety stock plays a central role in MRP so as to achieve an effective and stable scheduling. This matter is the most attracting yet challenging to researchers due to its highly necessity and importance to almost all managers, especially in manufacturing field. Along with that, the concern about managing safety stock in terms of dimensioning, positioning and replenishing is also examined carefully. It is quite understandable as manager not only cares about how many or how much stock should be kept but also how to manage them, e.g. storage space or timing.
Objectives
The last method to categorize topics on safety stock is according to their objectives since a paper may cover many issues involving safety stock. For instance, to find out the optimal of safety stock level, authors analyzed demand forecast, lead time, inventory policies as well as stock base. Nevertheless, once gain owing to huge number of issues regarding to safety stock, each paper was only successful in achieving certain number of objectives as one of the following objectives:
- Managing safety stock in terms of dimensioning, positioning and replenishing
- Using safety stock to improve MRP and MPS performance
- Optimizing safety stock level
Discussion
It is worth noticing that safety stock is different from inventory in a sense that it is a kind of strategic stock and most of research papers refer safety stock to finished goods inventory (exclude in-process inventory)
Since safety stock is a broad and important topic, there exist many issues concerning it and each research paper has managed to solve some of them. All of these papers will be discussed in this section but more analysis and efforts will be put for the subject of optimizing safety stock, managing them and study case of Kodak.
Relation between safety stock and lot-sizing and lead time
To begin with, the interrelation between safety stock and other factors namely lot-sizing and lead time was investigated carefully by some authors. The studies show that lot-sizing is an important issue in inventory management and the level of safety stock is definitely affected by lot-sizing. It is lot price and firm’s policy towards lot-sizing that define an appropriate safety stock level for a firm. As for this subject, article “The Uncapacitated Lot-sizing with Sales and Safety Stock” has presented a model to determine safety stock level more accurately. Together with lot-sizing, lead time also critically affects the determination of safety stock level. In effect, there are two areas that managers can improve to reduce inventory without hurting service level are reduction of lead time from suppliers and the variability of this lead time. The work of Eppen and Martin (1988) concluded that decreasing in both areas can reduce inventory. However, Chopra, Reinhardt and Dada through their research claimed that these conclusions which based on normal approximation are flawed, especially in the range of service levels where most companies operate. The authors proposed the existence of service level threshold greater than 50% below which reorder point increases with the decrease of lead time variability. Thus, for firm operating below this threshold but greater than 50%, reduction in lead time would lead to decrease in reorder point while decreasing in lead time variability would increase reorder point. Therefore, decrease in lead time is the right lever to cut safety stock, not reducing lead time variability. Moreover, graphs of safety stock as a function of lead time and lead time as a function of safety stock are drawn in order to capture the relation between them. The study also shows that since most firms are operating in the range of cycle service level from 50% to 60%, managers should better off reduce lead time instead of lead time variability to achieve the goal of reducing safety stock.
Optimizing safety stock
Graves and Willems in their paper, “Optimizing Strategic Safety Placement in Supply Chains”, developed a model to determine the safety stock at a particular stage of manufacture.
E[Ij] = Bj – E[dj(t – Sij – Tj, t – Sj)]
This equation with its variables and abbreviation simply represents that safety stock at stage j depends on the net replenishment time and the demand bound. This is for single-stage model.
As for the multi-stage model, safety stock level is also figured out by a more complicated formula. However, the noticeable point of these formulas is their applicability in situations that satisfy their assumptions. The key assumptions are: each stage of the supply chain operates with a periodic-review, base-stock policy; each stage quotes a guaranteed service time to its customers and demand is bounded. The feasibility of these models will be tested in the case of Eastman Kodak.
Eastman Kodak case
Study case of Kodak, a manufacturer of digital cameral, is a simple yet powerful in helping understanding as well as examining the feasibility of safety stock model built by Graves and Willems from MIT.
Since1995 Kodak has applied the model for more than eleven finished products across two of its assembly sites within its equipment division. The model’s application to the internal supply chain for a high-end digital camera will be elaborated here.
Product background:
Digital camera requires the following parts:
- A 35mm camera – procured from outside vendor
- An imager – produced internally
- A circuit-board – produced internally
- Other parts
The initial goal was to optimize the safety stock levels for the stages that were under the direct control of the final assembly area. The decision to start with the final assembly area, according to the study was based on the product’s high material cost and its relatively simple supply chain structure.
Manufacture and distribution process:
To produce digital camera, the back of the 35mm camera is removed and replaced with a housing containing the imager and circuit board. The camera is then tested. If the camera passes the quality test, it is shipped to distribution center. From the distribution center, the camera is shipped to customers
Model applied and result
-
Optimal solution: hold safety stock of components, subassemblies and completed cameras at the manufacturing site but no inventory at the distribution center
Total annual holding cost: $78,000
- However, the product flow team explored some near-optimal solution because they felt that some additional organizational constraints are not captured in the model
- Therefore, the real optimal: the distribution center holds safety stock but the manufacturing site would not hold (as it is shown in the figure above)
Total annual holding cost: $81,000
Conclusion
This case clearly illustrates the practicability of safety stock model. More essentially, it pointed out a significant fact that theoretical model, when applied in real-world has to take into consideration individual firm’s policies and constraints in order to get the best result.
Managing safety stock
From the above sections, safety stock level somehow can be calculated (though it remains many limitations), the problem of how to manage them arises. As a consequence, some of the papers focus on this issue and provide recommended guidelines of where to position, how to dimension and when to replenish safety stock. The study suggests that safety stocks kept to face demand uncertainty should be positioned on pull managed items, which is defined as items whose consumption is connected in a straightforward manner, i.e. pegged, to a forecast demand and does not refer to a stock replenishment. Positioning safety stocks on push managed items can reveal itself to be expensive for companies either because these items are customized or they are hard to keep (e.g. large-size or fragile products, perishable goods, etc). Hence, safety stocks are to be placed on items in pull with the market and this allows the manufacturing system to react on time whenever a forecast error occurs. As for dimensioning, the study shows that safety stock are proportional to the cumulative lead time to avoid stock-out on the lower-level push managed items. Interestingly, there is trade-off in replenishing safety stock. On the one hand, replenishment orders can be released as soon as safety stocks are withdrawn, so as to guarantee the appropriate service level of the company, but this introduces nervousness into the production plan. On the other hand, safety stocks can never be replenished so as to smooth down MRP nervousness, but this runs a high risk of stock-out. It is obvious that the later case does not fit real-life, however, the key point of these two cases lie in the combination of two scenarios to fit the firm’s policy as well as optimum safety. In short, all major aspects of safety stock have been presented and they appear to be useful for managers in coping with uncertainty in MRP environments.
The importance of safety stock on MRP performance measures
Most of the research papers agreed on the fact that safety stock reduces the adverse impact of forecast inaccuracy on the production schedule, because extra items are available to respond to unexpected increases in demand. More significantly, in a multi-level product structure managers should keep safety stock at the end-item level. If safety stock “absorbs” forecast inaccuracy at the top level of the product structure, it will prevent such inaccuracy form affecting the schedule for the lower level items as well. However, the magnitude or the size of safety stock that affects MRP performance. A high stock level could improve the stability of the MRP without degrading customer service but incurs high inventory holding cost. Low stock level increases the nervousness in production plan and sometimes cannot response quickly to the changes in demand yet it incurs low inventory holding cost. As a result, safety stock plays a vital role in evaluating MRP performance in terms of customer service, total cost and schedule stability.
Evaluating safety stock methods in MRP system
One of the articles on safety stock has discussed an attention-grabbing and interesting issue of how different methods of establishing safety stock affect MRP system performance. In fact, many authors have investigated the impact of safety stock on the performance of MRP system (as discussed in the previous section) but not many have investigated alternative methods of establishing safety stock in MRP systems. Since there are several ways of setting safety stock, it would be useful to evaluate how these methods. The result of this research shows that since various methods (in this study, 3 methods were examined) require different parameters or different method of calculating the same parameters, they produce different safety stock level. One method significantly influences total cost while others may significantly influence schedule instability or service level.
Remained and unsolved problems
Some authors mentioned unsolved problems in their papers, mainly problems with the assumptions they made in order to facilitate process of analyzing particular problems. To name a few:
- Assuming stochastic lead time, non- stationary demand, and limited capacity constraints. These assumptions are not realistic in the real-world since they are too simplified.
- Ignoring the impact of forecast error.
- Ignoring the combined impact of more than two factors at a time.
Comments on these research papers
All papers on safety stock have their own values to some extend. They are all trying to look into one among wide variety of problems with regards to safety stock. The results from these studies were successfully applied in the real world at certain conditions and some of them helped to orient future researches. However, most of the studies focused on figuring out safety stock for end-item level, not for lower level items. Low level items are usually not required safety stock under MRP system; nevertheless there are many contingent factors that call for the need of using safety stock. It is the case of late delivery or fabrication or assembly times are longer than expected or a bottleneck process. As a consequence, the use of safety stock for low level items is necessary to maintain smooth operations. In addition, with the aim of achieving research results with wider application, assumptions should be made to be more realistic and practical. Moreover, studying on safety stock should not be too model or mathematical-oriented, which is useful in theoretical analysis yet usually complicated, rather they should concentrate on helping managers solve three big problems:
- Which items, either finished products or components, should be endowed with safety stock?
- How should the buffers be sized?
- When should a replenishment order for safety stocks be released?
What’s more, it may be valuable and innovative to investigate the required safety stock in JIS system. This would be of help for manager since safety stock in JIT may be different from other systems because JIT is very “lean” system with very little waste. Hence, this could open a room for new research.
Conclusion
Literature reviewing is truly helpful for me to obtain better understanding on the topic I am interested in, safety stock. Through these research papers, I have gained valuable knowledge beyond the textbook about all the uncertainties and difficulties inherent with safety stock. Some papers successfully developed models to determine safety stock level and use examples of real companies to illustrate their practicality, which turns out to be very impressive to me. I did realize that all the issues discussed in these papers are very close to what we have been learning but the worth-mentioning point through this literature review is that authors have helped me to find out the links between discrete topics and safety stock, their interrelation and interaction. This is extremely important for any students so as to understand a problem more thoroughly and profoundly. Yet I faced problems in perceiving complex models (safety stock model, base stock model, etc) in these researches and some papers approaching the topic with mathematical modeling is harder to comprehend than other methods such as simulation modeling. In conclusion, though there remain some difficulties in understanding these research work on safety stock exhaustively, it is offset by providing me a broader and deeper view in dealing with safety stock.
Reference:
-
Stephen C. Graves & Sean P. Willem, “Optimizing Strategic Safety Stock Placement in Supply Chains”, Manufacturing & Service Operations Management, Vol. 2, No. 1, Winter 2000, pp. 68-83.
-
M Caridi & Cigolini, “Managing Safety and Strategic Stock to Improve Materials Requirements Planning Performance”, Journal of Engineering Manufacture, Vol. 216, Part B, 2002.
-
Xiande Zhao, Fujun Lai & T.S. Lee, “Evaluation of Safety Stock Methods in Multilevel Material Requirements Planning Systems”, Production Planning & Control, Vol. 12, No. 8, 2001, pp. 794-803.
-
Xue Bai, J. Steve Davis, John J. Kanet, Steve Cantrell & J. Wayne Patterson, “Schedule Instability, Service Level and Cost in Material Requirements Planning System”, International Journal of Production Research, Vol. 40, No. 7, 2002, pp. 1725-1758.
5. Sunil Chopra, Gilles Reinhardt & Maqbool Dada, “The Effect of Lead Time Uncertainty on Safety Stock”, Decision Sciences, Vol. 35, No. 1, Winter 2004.
6. Marko Loparic, Yves Pochet & Laurence A. Wolsey, “The Uncapacitated Lot-sizing with Sales and Safety Stock”, Online version, 2002.
-
S. C. L. Koh, S. M. Saad & M. H. Jones, “Uncertainty under MRP-planned Manufacture: Review and Categorization”, International Journal of Production Research, Vol. 40, No. 10, 2002, pp. 2399-2421.