THE PARADIGM OF LEAN RESUPPPLY: A TRAP IN THE MANAGEMENT OF THE FLOW OF MATERIALS BETWEEN LINKS IN THE SUPPLY CHAIN

This article presents a case study conducted in an industrial company, in which the logic of resupplying materials between its Distribution Center and its commercial branches was strongly influenced not only by targets for reducing inventory levels, but also by the perception that lean resupply (Just in Time) is the state of practice to be adopted by all companies. The company's operation, however, pointed to a growing increase in transport costs and in the frequency of stock-outs in its branches.

Through a spreadsheet simulation of alternative supply policies, it was assessed that lean resupply was not the lowest total cost logic, leading the company to initiate change management by understanding the origins and operational implications of the Just in Time philosophy. Furthermore, it was necessary to understand that the fragmentation of the logistic process by traditional accounting does not allow evaluating the existing trade-offs between inventory and transport costs. In this sense, the article addresses the concepts of lean resupply and total cost analysis of the supply chain in an integrated manner, as well as presenting a methodology for determining and adopting the most appropriate resupply logic for the cost structure.

1. INTRODUCTION

Currently, there is a large amount of space devoted in books, periodicals and specialized magazines to the principles of the Just in Time resupply philosophy (also known as lean resupply), as well as to the countless success cases that arise each year, reporting as the main result the dramatic reductions on average inventory levels. On the other hand, new systems for assessing financial performance, also with recognized prominence in the specialized literature such as EVA (Economic Value Added or Added Economic Value), advocate the continuous reduction of Net Working Capital (Working Capital) as a way to maintain competitiveness of the company in the long term. The mass diffusion of these new trends in the business environment can create some biases in the management of the flow of materials in the supply chain. This is the main theme that we intend to explore in this article, when we present a case study of a company in the industrial sector.

In this sense, in section 2 we discuss the lean resupply paradigm, its origins and its implications on the trade-off between inventory and transport costs, questioning whether there is any business segment in which the savings obtained with the reduction of average levels of stock would not compensate for the increase in costs resulting from a higher frequency of transport pulses. Section 3 presents the company, its insertion in the supply chain, its main products, the guidelines that govern the current logic of activating the flow of materials and the main problems currently verified, such as, for example, the explosive increase in expenses with transport and the frequent stock-outs of products. Section 4 comments on the results obtained with the Excel spreadsheet simulation of alternative transport policies, interpreting them in the light of the economic lot approach, as well as other relevant concepts in the business logistics area, such as total cost analysis and suitability of lean resupply restricted to some specific situations. Finally, section 5 concludes this article, presenting a six-step methodology that encompasses all the ideas discussed throughout the text and aims to help companies assess their current materials management, vis-à-vis the stock/stock trade-off. transportation and consistency with performance measurement systems.

2. THE LEAN RESUPPLY PARADIGM: ORIGINS AND CONCEPTS

Originally, the lean production regime conceived today as Just in Time encompassed a set of various manufacturing, purchasing and distribution procedures adopted by Toyota to restructure its production system in the late 40s. The basic objective in adopting these procedures was to produce cars at the lowest possible cost, making them more competitive than US counterparts. The team of production engineers led by Taiichi Ohno, when perfecting one of these procedures, the mold changing process in sheet steel pressing, found that the cost per piece pressed was lower in small batch production than in batch processing. immense.

According to WOMACK et al. (1992), there are two basic reasons for this phenomenon. The first is that the production of small batches eliminated the very high opportunity costs of capital in maintaining inventories of finished parts, compared to North American systems of mass production. In fact, by reducing mold changeover time from one day to three minutes, Ohno's team has shifted the balance of the trade-off between the cost of batch processing (basically fixed labor costs and downtime allocated to each preparation) and the cost of maintaining inventories to a lower level, as shown in figure 1. On the other hand, the production of small batches made production errors more visible, motivating, even if not premeditated, a crusade against waste in production and against defective parts.

This example illustrates why several authors on lean production regimes (SLACK, 1993 and CORREA et al., 1994) classify the economic batch production approach as “traditional” and “reactive”. The classic equation for calculating the economic lot size is described below, where D is the annual demand, CP is the processing cost associated with a single lot, and CME is the annual cost of holding a unit in stock (sum of the opportunity components of capital, space, obsolescence and perishability).

This approach, in addition to being limited because it does not consider possible variability in demand and processing lead-times (SILVER et al., 1985), can favor decision-making in which the maintenance of the efficiency frontier of the production system is perceived as immutable . This is because the acceptance of the parameters used in the equation can inhibit questioning, review and continuous reduction of operation costs (CORREA et al., 1994) through initiatives similar to those adopted by Ohno and his team.

The continuous improvement philosophy has also extended to other areas such as, for example, the purchasing process and the management of suppliers in the supply chain. WOMACK et al. (1992) addresses how Toyota developed a unique way to coordinate the flow of parts in the supply chain through Kanban cards, resulting in Just in Time resupply. These cards triggered the consolidated transport of parts, usually in small containers, from suppliers to the company whenever necessary. It should be noted that, in this specific case, the increase in transportation expenses (due to a higher frequency of pulses) was more than offset by the reduction in the opportunity cost of maintaining inventories in the supply chain, leading the system to a more efficient operation. lowest total cost, as shown in figure 2.

As noted by BOWERSOX et al. (1996), the diffusion of technologies for the electronic exchange of data between companies, allowed that several Just in Time resupply procedures originated in the automobile industry to be structured in other supply chains, mainly in the food and clothing chains from the 80s onwards. In this regard, we highlight the Quick Response Program (QR) in clothing retail and the ECR Movement (Efficient Consumer Response) in food retail.

In the specific case of these programs, the increase in transportation expenses (due to a greater number of trips between manufacturers and retailers) was more than offset by reductions in the total cost of inventories. Although the added value per unit weight of clothing and most food products is comparatively low to that of parts and components in the automobile industry (and consequently the opportunity cost of capital), two other components of retail stand out. cost of holding inventories: the cost of space occupied and the cost of obsolescence and/or perishability of products. The escalation of prices per square meter in several urban centers around the world, associated with the progressive shortening of the life cycle of several products, have led several retail chains to adopt automatic resupply regimes with their main suppliers. Figure 3 illustrates this issue.

The question that should be asked at this point is whether there is any business segment in which reductions in the size of shipments would not compensate for any diseconomies of scale in transportation. In other words, the aim is to identify and characterize for which business segments transport costs are significantly higher than inventory maintenance costs, making lean resupply regimes economically unfeasible. The case presented in the next section fits perfectly into this cost profile.

3. CASE STUDY

The analyzed company is an important producer and distributor of industrial and medical gases in Brazil, with gas production units spread across several states. The sale and distribution of gases accounted for most of its revenues in 1998. Approximately R$ 50 million of total revenues come from the sale of equipment such as valves, pressure regulators, torches, welding torch nozzles, etc. to building materials wholesalers. and hardware and for several small clients such as mechanical workshops and body shops. The company believes in the synergy between gases and these equipments when selling to these types of customers, with complementary products that reciprocally leverage sales.

More specifically, the distribution and commercialization of these types of equipment constituted the focus of this analysis. We will detail the company's supply chain below, as illustrated in figure 4.

The company has a distribution center located in the municipality of Rio de Janeiro, which is supplied by three main sources of supply: a cryogenic equipment factory, also located in the municipality of Rio de Janeiro, several small manufacturers located in the state of São Paulo and international suppliers. The three sources account respectively for 80%, 10% and 10% of the almost 3.000 tons of equipment purchased in 1998.

The distribution center, in turn, supplies more than 50 commercial branches throughout Brazil, in addition to selling directly to large wholesalers, a sales modality that represents seventy percent of total revenue from equipment. In turn, the commercial branches are intended for the direct service of small industrial customers and mechanical workshops.

The procedures for triggering the flow of materials in this supply chain differ substantially from the Supplier-DC link to the DC-Branch link. The supplier-DC link follows the strict logic of monthly purchase schedules, in which at the end of each month the commercial operations management sends purchase intentions per product for the next four months to its suppliers. The values ​​for the first month are unchanged, for the second and third months changes of +/- 15% are allowed and the fourth month remains open. The resupply lead time on the Supplier-DC link varies from 30 to 40 days and suppliers generally do not accept rescheduling for the first month or emergency shipments.

The DC-Branch link follows a more flexible resupply logic, whereby transport pulses are triggered whenever the stock level per product (physical and in transit) at each branch falls below half of the Maximum Inventory Threshold (PME). The principle of operation is similar to a two-drawer system (WATERS, 1992), ensuring that the volume of products in stock will never exceed a maximum ceiling, in this case the SME itself. Figure 5 illustrates the simplified logic of the two-drawer policy.

These distinct logics for activating the flow of materials in each link entail significant consequences in the operation of the supply chain, in particular for the Distribution Center. Are they:

• The Distribution Center assumes the role of the system's buffer stock, as it seeks to ensure high product availability, given the high inflexibility in rescheduling orders with suppliers, to meet the demands of the branches.
• Demand from the Distribution Center cannot be analyzed as a dependent demand (or the sum of demands) from its branches for the purposes of purchasing programming. Applying the Distribution Resource Planning (DRP – Distribution Resource Planning) logic to the Distribution Center would imply redefining the type of relationship with its suppliers, with a view to reducing supply lead-times. A more detailed discussion on these aspects can be found in BOWERSOX et al. (1996).
• In the short term, improvements in the way purchases are scheduled for the DC include the adoption of more accurate forecasting methods. In this case, it was found that the application of the simple exponential smoothing method in place of the three-month moving average would bring increases of up to 80% in the level of accuracy of purchase schedules.

On the other hand, the DC-Branch link presents some inconsistencies in the definition of inventory resupply parameters. The PME parameter, for example, is calculated as the moving average of the daily sales peaks of the last three months. This calculation disregards:

• Different resupply lead times for branches
• Variability in demand and resupply lead times for branches
• Desired service levels, measured in terms of average product availability on branch shelves.

This methodology for calculating the maximum inventory level per product (PME) has significant consequences for the operations of the branches. Are they:

• Very frequent resupply of small quantities. One of the reasons why the PME has a relatively low inventory level and that it goes up very frequently is the fact that it does not incorporate the transport lead-time into its calculation, as well as the demand variability in this lead-time.
• Extremely unconsolidated shipments. In 1998, 80% of shipments shipped from the Rio Distribution Center weighed less than 100 kg. Due to the fact that these pieces of equipment have low added value (about R$ 50,00 to R$ 100,00 per unit) and weight (1,5 kg per unit on average), the deconsolidation of loads excessively costs the sending of several pulses a month.
• Low product availability on branch shelves due to mismatched batch sizes shipped with resupply lead-times and variability in transport time and demand. Only 40% to 50% of customer demand was met immediately.

The Commercial Management and the Company's Branches, however, considered the resupply policy parameterized by the PME to be extremely adequate for its operations. Among some of the reasons given, we highlight:

• Frequently shipping small batch sizes supports the Just in Time resupply philosophy described in the previous section. From the perspective of its managers, the company was operating according to the most modern state of operations management practices. Some branches even further advocated reducing the size of the shipment lot and increasing the frequency of resupply.
• The main guideline for measuring the company's result was the Reduction of Working Capital, denomination in English for Net Working Capital. Within this perspective, each business unit, such as the Distribution Center in Rio and the commercial branches, should adapt to the targets for the progressive reduction of Accounts Payable, Accounts Receivable and, above all, Inventories. From the perspective of its managers, the policy of resupplying in small batches was perfectly suited to the Working Capital reduction goals adopted since the acquisition of the company by an American counterpart three years ago.

We clearly perceive that the influence of the zero inventory paradigm created from the diffusion of several success stories in the automobile industry and retail (as explained in the previous section), associated with economic-financial performance evaluation criteria that do not fully consider the trade -off between the costs of maintaining inventories and transportation expenses, contributed to the company's management not taking an adequate position regarding the best resupply policy for the DC-Branch link. In fact, the accounting breakdown of various costs of the logistical process, as illustrated by figure 6, in which distribution expenses are posted on the Income Statement for the year and inventories constitute "activated costs" on the Balance Sheet, can prevent managers from Note two basic aspects:

• that the option for lean resupply or Just in Time, although there are numerous success stories, is fundamentally explained by the analysis of the stock/transport trade-off, and constituted, in its beginnings, a particular solution to the specific problem of Toyota (PROENÇA , 1994).
• that the stock/transport trade-off and the need for the analysis of total resupply costs, although not refined, elaborated and subject to various criticisms (SLACK, 1997 and CORREA et al., 1994), had already been addressed in the methodology of economic lot proposed by Wilson in the 1930s (WATERS, 1992).

The following section describes the reassessment of the resupply policy in the CD-Branch link based on a simulation of the operation in an electronic spreadsheet. The results obtained are discussed in light of the assumptions of the economic lot model, as well as the existing theoretical framework in the areas of logistics and operations strategy.

4. EVALUATION OF THE COMPANY'S RESUPPLY POLICY VIA SIMULATION: DISCUSSION OF THE RESULTS OBTAINED

Several authors in the field of operational research, such as WATERS (1992), TAVARES et al. (1996) and RAGSDALE (1998), defend the use of electronic spreadsheets as a tool to support the formulation and choice of resupply policies, even though there are analytical tools available for this purpose. The flexible nature of spreadsheets allows you to easily:

• Evaluate variations in the demand profile of products over reorder points and safety stocks,
• Evaluate variations in the opportunity cost of capital on the parameters of the operation,
• Evaluate how variability in replenishment lead times affects reorder points and safety stocks,
• Evaluate how variations in the rate of obsolescence and perishability of products affect policy parameters,
• Simultaneously consider the impact of unit quantity discounts and freight cost discounts by weight range on the size of purchase lots,
• Consider the component of the opportunity cost of inventory in transit over the total cost of the operation,
• Collect simulated statistics on service level (product availability on the shelf) and total cost of operation,
• Evaluate the impact of dimensioning safety stocks, based on the desired probability of stock-out per cycle, on service levels.

In addition to making it possible to assess these considerations relevant to a single link in the supply chain (in this case, the DC-Branch link of the company under study), SILVER et al. (1985) point out several other considerations pertinent to more than one link that can be easily simulated in an electronic spreadsheet. Are they:

• Evaluate the impact of stock-outs at the Distribution Center on the availability of products at the branch,
• Determine the best apportionment policy to meet the demands of the branches when there is not enough stock in the Distribution Center,
• Evaluate the impact of alternative modes of transport (eg premium transport) on operating parameters.

With all these considerations in mind, the current DC-Branch resupply policy, determined by the PME, and alternative reorder point and lot size policies were simulated for a period of 365 days. Figure 7 represents a very characteristic example of the results found in terms of the total cost of the operation, consisting of the comparison between the current policy and the alternative for a pressure regulator of R$ 50,00 (cost of acquisition with ICMS included) sent to the Northeast Branch I from the DC in Rio de Janeiro (consolidated freight cost of BRL 24,24 for a weight range of up to 100 kg). Other relevant information:

• The daily demand for this product can be approximated by a Normal probability distribution (0,90 ; 1,55)
• The opportunity cost of capital was assumed to be 24% per year
• The degree of perishability and obsolescence of this product is negligible.

We can see clearly from figure 7 that the current policy (PME), according to reasons already explained above, points to a resupply batch size of 3 units, 116 shipments per year and total cost of the operation of R$ 2.500,00/year. The alternative policy consists of a batch size of 37 units, 8 shipments per year and a total cost of operation of approximately R$ 500,00/year. The average product availability on the shelf increased from 47% to 90%.

These results are in line with the conceptual modeling implicit in the economic lot size formula:

• for products with low added value, the lower the opportunity cost of maintaining inventories, consequently the greater the propensity to ship larger batches;
• for products with relatively little demand, transport costs become proportionally higher compared to the value of the transported cargo. In this sense, there is also a greater propensity to send larger batches.

Even though the Working Capital at the Northeast Branch I has increased, due to the average inventory level having gone from 1,5 units to 18,5 units, the total costs of the operation were reduced. According to LAMBERT et al. (1998), it is necessary for business managers to have a systemic view of commercialization (marketing) and distribution (integrated logistics) objectives in order to minimize total costs, instead of seeking to reduce each component of these costs in isolation, ignoring each other existing trade-offs. For a given configuration of commercial decisions regarding product, price, promotion and place/customer service (the 4 P's of the Marketing Mix), one must seek to design the logistics operation that results in the lowest sum of transport, inventory, storage, processing costs of orders, etc. Figure 8, adapted from LAMBERT et al. (1998) illustrates this perspective.

Another critic of the economic batch approach as “traditional” and “reactive”, because it does not indicate how to shift the system efficiency frontier through the continuous questioning of operations, CHRISTOPHER (1997) agrees that for certain product and demand characteristics, the lean resupply approach is not the most appropriate. As the philosophy of Just in Time resupply is to deliver small quantities to their destination, more frequently and at the exact time they are needed, the challenge for supply chain management is to find ways in which these requirements can be satisfied, without increasing undesirable costs. There may be trade-offs, but the ultimate goal should be to reduce total supply chain costs.

In this sense, the lean resupply option may not always be adequate or justified for components with low added value, or in situations where demand is relatively predictable. In the automobile industry, the situation is exactly the opposite of the company under study: components with high added value and little predictability in the demand for components, since there are several possible variations of style, shapes and colors for a single category of parts (eg . bumper). Figure 9 illustrates in which situations it is most advantageous, in terms of total supply chain cost, adoption of lean supply regimes or variations of the economic lot approach.

We have seen so far that the lean resupply paradigm associated with the way in which the result of a business's operations is evaluated, in this case the reduction of Working Capital, can lead a company to mistakenly position itself in the management of the flow of materials between the links of the supply chain. In this sense, change management was necessary to establish consistency between goals and operations, in addition to convincing people that change was really necessary.

The company is currently reformulating its performance indicators to include the total costs of the supply chain, that is, it is incorporating transport costs to inventory costs at the Distribution Center and at the Branches. In addition, customer service indicators are being measured, such as product availability on shelves. Within this context, the increase in inventory levels is perceived as an investment that will provide not only direct returns, such as a reduction in inventory costs, but also indirect returns, such as a reduction in the level of lost sales and back-orders. from increased product availability. Figure 10 illustrates the performance pyramid that is guiding the change management process.

In view of all these considerations, at the conclusion of this article, we present some basic steps that must be followed by companies that wish to assess adequacy and, if applicable, reposition the logic of managing the flow of materials in the supply chain.

5. CONCLUSION: METHODOLOGY FOR DETERMINING THE MOST APPROPRIATE RESUPPLY LOGIC

The analysis of material flow management between links in the supply chain basically goes through six main stages, described below.

Stage 1: Initial sensitivity analysis using the economic purchase lot approach

It allows assessing whether the lot size currently practiced by the company differs significantly from the economic lot, providing subsidies on the relative importance between the cost of maintaining inventories and the cost of processing orders in the current frontier of system efficiency. Because the economic lot approach is robust to lot size variations (as illustrated in figure 11), significant short-term changes in operation should only be taken into account if the relative difference exceeds 50%.

 

Stage 2: Demand variability analysis

It is a significant component in determining the most appropriate resupply regime. Products with a high Average Absolute Deviation (error) in the sales forecast tend to have high safety stocks, increasing the cost component of maintaining stocks and making the option for lean resupply more interesting. It is easy to prove that for each unit increase in DMA, safety stocks (ES) increase by 2,45 units considering a 95% confidence level that there are no stockouts. The demo follows:

 

Stage 3: Simulation of alternative inventory policies

As previously mentioned, the simulation allows the incorporation of several specific characteristics of the analyzed operation without the need to use or elaborate more elaborate analytical formulations. Details such as quantity discounts, freight costs varying by weight range, contribution margins lost due to stock-outs can be easily incorporated into the simulation, leading to more accurate determination of lot sizes and operating costs.

Stage 4: Analysis of results and generation of scenarios

Once the current resupply logic, its costs and operating parameters are validated, different scenarios can be evaluated that imply a substantial change in the efficiency frontier of the system and shift the size of the order lot. For example:

• hiring a logistics service provider responsible for consolidating the volume to be transported by the company with shipments from other shippers, resulting in a substantial reduction in the cost of processing orders,
• the formation of partnerships with customers, in which the exchange of sales information in real time, via EDI, allows the elaboration of more precise production and distribution schedules,
• the heightened competition, leading to substantial increases in the service levels expected by customers.

It should be noted that the simulation of resupply policies makes it possible to estimate, for each new scenario evaluated, the new total costs of the operation, serving as a starting point for estimating the economic and financial viability of the change in relation to the current scenario.

Stage 5: Evaluating the company's recognition and reward policy

It is a fundamental factor for defining the change management process. As seen throughout this article, the definition of the objective to be achieved by the company, in this case the reduction of Working Capital, can create a bias in choosing the most appropriate resupply policy. At this stage, the congruence of the company's recognition and reward systems with the analysis of the results (costs and operation parameters) obtained via simulation should be evaluated.

Stage 6: Adoption of integrated performance, cost of operation and service level indicators

This step is an immediate consequence of the assessment made in the previous step. Generally, an attempt is made to establish an association between performance indicators that reflect the level of service provided and the total costs of the operation incurred, as opposed to a fragmented perspective of the logistics process.

BIBLIOGRAPHY

BOWERSOX, DJ, CLOSS, DJ 1996, Logistical Management – ​​The Integrated Supply Chain Process, 1st ed, McGraw-Hill.

CHRISTOPHER, M., 1997, Logistics and Supply Chain Management – ​​Strategies for Cost Reduction and Service Improvement, 1 ed. São Paulo, Pioneira Publishing House.

COPELAND, T., MORRIN, J., 1996, Valuation: Measuring and Managing the Value at Companies, 2nd ed. NewYork, J. Wiley.

CORRÊA, H., GIANESI, IGN, 1994, Just in Time, MRP II and OPT – Um Enfoque Estratégico, 2 ed. São Paulo, Editora Atlas.

LAMBERT, DM, STOCK, JR, ELLRAM, LM, 1998, Fundamentals of Logistics Management, 1st ed. New York, Irwin-McGraw Hill.

MENTZER, JT, BIENSTOCK, C., 1998, Sales Forecasting Management, 1st ed, New York, SAGE.

PROENÇA, A., 1994, “Production Management and Competitiveness: Assumptions of the Dynamic Capabilities Approach in Production Strategy”, D.Sc. Thesis, Rio de Janeiro, Production Engineering Program – COPPE/UFRJ.

RAGSDALE, CT, 1998, Spreadsheet Modeling and Decision Analysis – A Practical Introduction to Management Science, 2nd ed. Cincinati (Ohio), South Western College Publication.

SILVER, EA, PETERSON, R., 1985, Decision Systems for Inventory Management and Production Planning, 2nd ed, New York, Wiley & Sons.

SLACK, N., CHAMBERS, S., 1997, Production Management, 1st ed. São Paulo, Editora Atlas.

SLACK, N., 1993, Competitive Advantage in Manufacturing – Achieving Competitiveness in Industrial Operations, 1 ed. São Paulo, Editora Atlas.

TAVARES, LV, OLIVEIRA, RC, 1996, Operational Research, 1 ed, Alfragide, Portugal, McGraw-Hill.

WATERS, CDJ, 1992, Inventory Control and Management, 1st ed, New York, Wiley & Sons.

WOMACK, JP, JONES, DT, ROOS, D., 1992, The Machine that Changed the World, 1 ed., São Paulo, Editora Campus.