Description

This downloadable virtual product contains a SimPy simulation app model and library that is made accessible to end-users via a ShinyR app with a graphical user interface (GUI). The simulator simulates yogurt production, distribution, and sales and availability at end warehouses / consumption facilities (= supermarkets). The app is demonstrated in below Youtube video. It is a ShinyR app that uses a SimPy simulation library that I developed in Python for implementing a yogurt supply chain simulation model. If you purchase this product you will get access to the ShinyR app in the form of a R-file, and the underlying Python model and library – all in the form of a downloadable zip-file for execution on your laptop or PC. If you prefer, this app can also be made available as a web app, without you having to access the underlying source code. For this please contact me through our contact form here.

Assumed supply chain network structure

This model assumes a supply chain structure with one (1) central warehouse / manufacturer, and three (3) decentralized warehouses (= supermarkets).

The manufacturer produces according to a defined production plan and program. The manufacturer produces onto stock, which is the centralized warehouse from where all decentralized warehouses are supplied once they place a purchase order at the manufacturer. That is, decentralized warehouses are supplied from stock, and the manufacturer’s inventory is managed according to FIFO principles. Once supermarkets, i.e. the decentralized warehouses, place a purchase order at the manufacturer, the product is shipped if available – and arrives after a defined delivery lead time.

Consumers arrive at the decentralized warehouses (= supermarkets) and buy yogurt from stock. They have defined preferences and the supermarket sells yogurt according to FIFO principle.

This model, in its default version, does not implement backlog. A consumer sales order pr supermarket purchase order at the manufacturer is either successful or unsuccessful.

Assumed consumer behaviour and product lines

There are three assumed yogurt product lines:

• strawberry
• blueberry
• apple

Consumers have defined preferences, and purchase what is available at the supermarket in accordance with their preferences:

• Group 1: Only want strawberry yogurt
• Group 2: Prefer strawberry, but accept blueberry
• Group 3: Only want blueberry yogurt
• Group 4: Prefer blueberry, but accept strawberry
• Group 5: Only want apple yogurt

Model parameters (product, consumer, production and distribution)

The yogurt supply chain simulator app allows users to adjust the following parameters (and to re-run the simulation with those updated parameter values to see the updated results):

• Simulation duration [days]
• Product validity [days]
• Mean daily production output at manufacturer [un]:
• Deviation in daily production output [un]:
• Whether a gaussian normal distribution (TRUE) or a log-normal distribution (FALSE) is used for modelling daily production output
• Share of strawberry, blueberry, and apple product lines of total daily production output
• Mean daily total yogurt consumer demand, for all end-warehouses (supermarkets), with associated deviation [un]
• Whether a gaussian normal distribution (TRUE) or a log-normal distribution (FALSE) is used for modelling daily consumer demand
• Share of consumers per preference: only strawberry, strawberry or blueberry, only blueberry, blueberry or strawberry, and only apple
• Mean daily total yogurt repurchasing volume per warehouse (= supermarket), and associated deviation [un]
• Delivery lead time [days] for yogurt ordered by warehouse (= supermarket) at manufacturer
• Number of warehouses (= supermarkets)

If manufacturing output is following a gaussian normal distribution, warehouse repurchasing volume follow this distribution type too. And vice versa, if manufacturing output is log-normally distributed, warehouse repurchasing volume is too. These two distribution types are linked.

Possible extensions and customizations

The app, and underlying simulation library and model, can be customized and tailored to other supply chain networks and logics. Here are some exemplary adjustments and customizations that I can implement into the app for you, in the form of a follow-up consultation:

• Deviating supply chain network, e.g. with additional network stages (cross-dock centers, distribution over multiple stages, etc.)
• Deviating inventory and ordering policies, e.g. order-point based inventory and purchasing policies, or forecast-based purchasing strategies
• Deviating production planning and control systems, e.g. make to order instead of production on stock
• Any number of warehouses, distribution centers, manufacturers
• Different consumer behaviours at different supermarkets (end-warehouses), and different demand patterns at differnt supermarkets (end-warehouses)
• Additional KPIs: Transportation costs, warehousing costs, production and scrapping costs, sales revenue, and more
• Pricing policies, e.g. regional pricing and price campaigns with resulting impact on inventory, availability, production volume, and profitability
• Consider backlog, at manufacturer(s) or if relevant also at the supermarkets

Check below video if you want to learn more about the app and the downloadable zip-file available with this virtual product.

Related content and products

If you are interested in supply chain simulation or production simulation solutions and content then the following products and articles are of interest to you:

• Link: SCM SimPy library with R interface
• Link: Job shop scheduling library in Python
• Link: Perishable product SCM SimPy library (Python)
• Link: Poultry end-to-end supply chain simulator
• Link: Warehouse assignment model in Python