Simulation for yogurt supply chain optimization

Dairy products are especially challenging when it comes to inventory management, production planning, warehousing, and distribution. One reason for this is the expiry date of products, and the need for cooling. A supply chain simulation app was developed and deployed for yogurt supply chain optimization, assessing changes to inventory, delivery, and warehousing policies. Relevant key performance indicators (KPIs) considered in this study covered availability, transportation costs, warehousing costs, sales revenue, and product loss expenses due to expiration date violations. The results were (a) an improved supply chain network design and (b) overall better availability with, at the same time, lower warehousing and product loss expenses.

A simplified version of the yogurt supply chain simulation app applied in this study can be found here in the SCDA shop:

Comprehensive customizations and additions were added to the model template linked above, including e.g.:

  • forecast-based production master scheduling
  • shipping and distribution strategies
  • demand-inventory policies (including known regional price campaigns)

Other adjustments included e.g. milk-run deliveries from decentralized end-warehouses to small supermarkets and stores.

Simulated yogurt supply chain distribution network

Below figure shows the relevant part of the simulated distribution network for yogurt production and distribution in Brasil. The simulated network is a subset of the overall distribution network. Distribution, sales, warehousing, and transport was only simulated at and between these highlighted facilities and supply chain entities.

Distribution network considered by yogurt supply chain optimization

The inbound milk supply, supplying yogurt production with milk, was not within the scope of this study. The relevant supply chain entities were the yogurt manufacturing plants, the supermarkets, and lastly smaller stores. Both supermarkets and small stores are supplied by a milk-run transport strategy, but the vehicles used for supermarkets have a larger capacity than those used for small stores – which is the main criteria for distinguishing small stores from larger supermarkets.

Management also used the simulator app to assess the option of introducing intermediary distribution centers, as they e.g. reduce transportation costs and can help in increasing inventory control throughout the entire supply chain. For example, supermarkets can be resupplied multiple times daily from an intermediary distribution center in the center of the city. In this way supermarkets are always supplied with small batches of yogurt, all with the same expiration date. This helps the manufacturer in maintaining control of product consumption, ensuring supply-chain wide First-In-First-Out principle adherence.

Product groups and consumer behavior

From a marketing point of view, consumers have preferences. The manufacturer produces 8 different yogurt types, all in the same size but with different tastes:

  • yogurt with honey
  • yogurt with strawberry
  • yogurt with blueberry
  • yogurt with coconut
  • yogurt with apple
  • yogurt with peach
  • yogurt with raspberry
  • yogurt with blackberry

Based on information from the marketing team, various consumer groups were modelled. For example, some consumers will like apple more than raspberry, but will buy apple yogurt if no raspberry yogurt is available. Other consumers might only want one specific type of yogurt and will not buy other yogurt types if the preferred type is not available. Or, for some consumers, if the preferred yogurt type is not available, they will still buy other yogurt types, but in a smaller amount than what would otherwise have been the case. For this reason, the simulation comprises sales revenue as one of the KPIs. The sales revenue refers to the sales revenue in the supermarkets and stores.

The supply chain simulator allows for analysis of complex and competing effects

The supply chain simulator allows decision makers to run what-if scenarios, assessing potential management decisions without running the risk of negatively influencing the live system. The simulation study considered the effects discussed in this section of the article.

Higher transportation frequency increases transportation costs. However, it also reduces average inventory levels at the supermarkets and thus in the entire supply chain – while at the same time also improving availability. This means that an increase in transportation frequency, e.g. once per day instead of every second day, also increases customer satisfaction and sales revenue. It also reduces product losses. That is, profitability can both benefit from and suffer under increases in transportation frequency. A simulation can run many different scenarios, thereby supporting management in their search for a cost optimum.

Improved cooling equipment and cooling chains increase capital expenses and variable energy expenses, but they increase product validity and postpone product expiration dates, improving quality and consumer satisfaction, as well as availability and sales revenue – in addition to reducing product loss expenditures.

Centralized vs decentralized inventory: Related to the transportation frequency dilemma, management has to decide what the optimal supply-chain wide inventory policy should look like. For example, yogurt can be stored centrally at the manufacturing plant, or at an additional intermediary distribution center in close proximity to the supermarkets. Having the main portion of the inventory at the distribution center or manufacturing plant will give the yogurt manufacturer more control over supply-chain wide inventory policies. In this case, a First-In-First-Out inventory policy was preferred, as this reduces the risk of product losses due to expiration date violations. With a larger centralized inventory, supermarkets and small stores would receive small batches from the distribution center or manufacturing plant, all with the same expiration date. In this way it can be avoided that consumers select yogurt with more distant expiration dates instead of yogurt with more near-term expiration dates. Product losses and supply-chain wide instabilities are thereby reduced, i.e. improved.

Intermediary distribution centers: Storing more yogurt at distribution centers would (considering the entire supply-chain) reduce expenses for property leases when compared to storing that same inventory at supermarkets and small stores. The reason for that is that the central warehouse lease per surface area is cheaper than the lease paid by most of the supermarkets. However, compared to storage at the manufacturing plant, a distribution center in most cases has to pay higher leases per surface area – which in that cases increases the costs. Furthermore, a distribution center generally reduces transportation costs, as products from the manufacturing plant can now be transported pallet-wise, i.e. in full truck loads (FTL). However, referring to above “centralized vs decentralized inventory” dilemma, the distribution center can be used for reducing inventory at the supermarkets by increasing resupply frequency from the distribution center. This will improve inventory control, reduce product losses and improve availability, sales revenue, and consumer satisfaction, but it also increases transportation costs.

Concluding remarks related to yogurt supply chain optimization

In this simulation study a Yogurt SCM simulator, available in the SCDA shop, was modified and tailored to a client-specific supply chain with 4 large yogurt factories and many supermarkets and small stores in Rio de Janeiro, Brasil. In this simulation, three major decision categories were considered: (1) Investments into improved cooling chains and cooling safety, (2) investments into an intermediary distribution center, and (3) larger centralized inventory levels for better inventory control and reduced product losses and resulting bullwhip effects.

Below video demonstrates the underlying SCM simulator app template used and customized for this simulation study. The simulator app was developed in SimPy and ShinyR. For this study, the simulator was adjusted and enhanced, as part of a customization effort.

You might also be interested in the following SCDA products and articles if you want to learn more about supply chain simulation:

You May Also Like

Leave a Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.