This simulation extends the basic "beer game" or supply chain model to demonstrate how different companies might react to exogenous supply and demand shocks. This model demonstrates how Toyota's stockpiling requirements have helped it weather the supply and demand shocks of the COVID-19 pandemic, in contrast to competing automotive companies with a just-in-time ordering strategy.

The Model

The model consists of a chip supplier, three firms, and 3 consumer demand agents. Each demand agent corresponds to, and generates demand for one of the three firms. The chip supplier produces chips at a certain variable rate and sends them to the three firms, and the three firms consume chips into sellable stock.

The Firms

Two of the firms in the model are automotive manufacturers, and one represents a consumer electronics firm. Both consume the chips from the chip_manufacturer.

In the 3D Viewer, firms are represented as both their produced stock and raw stock. Raw stock is represented by the height of the red bar, and produced stock by the height of the blue bar.

Consumer Demand

All three firms have a corresponding consumer demand agent which sends orders to the firms. These agents are affected by exogenous shocks which cause their weekly demand levels to change. When a demand agent receives a delivery, it records the fulfillment time (between when it placed the order) and records the average time.

Production

Firms can adjust their production rates gradually, based on the demand they are experiencing. Adjustments are made based off of a 5-week moving average of past orders. The model uses the greater of the average orders, or the average deliveries made. The production rate can change at a rate of 0.1% per week.

Delivery

Firms and the chip producer use messages to deliver stock. The stock the chip_producer delivers becomes the raw stock of the next firm down the chain (auto firm or consumer electronics firm). Firms and the chip producer attempt to fully or partially meet as many of their orders as possible.

The chip producer uses the expected price margin to prioritize which firms it will try and satisfy first. The profit margins are equivalent for the two auto firms, and doubled for the consumer electronics agent.

Ordering Strategies

Firms and consumer agents use messages to place their orders. The logic is captured in the producer_calculate_order.js behavior. All agents use the same ordering strategy except for the Toyota agent.

The basic strategy is "just-in-time", which means that agents will only try to stockpile enough raw material for one month worth of orders.

Since a series of natural disasters in 2011, Toyota requires 2-6 months of raw inventory, which means that the agent will place an order whenever the state.raw_stock falls below 8 * state.avg_delivery and will order enough to bring the level up to 26 * state.avg_delivery each time.

Recreating the Effects of COVID-19

We want to use this model to try and understand why Toyota's strategy is better in the face of the effects of COVID. To do this, we'll define exogenous shocks that correspond to some of the changes in demand and production seen during the pandemic.

Exogenous Shocks

The exogenous_shocks.js behavior is responsible for initiating changes in demand and production rates. The changes are encoded in the corresponding objects in globals.json.

The changes in demand and supply roughly simulate the effects of the COVID-19 pandemic, and its subsequent effects on the auto industry, chip production, and the consumer electronics market. When the pandemic begins around step 500, the demand for cars drops, and the demand for consumer electronics jumps up. Around step 1000, due to easing restrictions and a "return-to-normal" mentality, car sales come back to their pre-pandemic levels, and consumer electronics demand remains high.

Understanding the Outputs

If we examine the very first plot in the Analysis View, we can see that the fulfillment time for Toyota remains constant throughout these shocks. The consumer demand industry sees increases in fulfillment time, even though it is being prioritized by the chip manufacturer. The manufacturer still simply cannot meet the demand, and this matches the shortages that we are still seeing currently with shortages of stock in laptops and gaming consoles.

The competing automotive company does even worse than the electronics industry, since it has neither a safe stockpiling strategy like Toyota does, or the benefit of prioritization by the chip manufacturer. This leads to very significant increaes in the fulfillment time, matching the real-world shortages we are seeing.