Supply Chain Models Jayendran Venkateswaran IEOR @ IIT Bombay - - PowerPoint PPT Presentation

Supply Chain Models

Jayendran Venkateswaran IEOR @ IIT Bombay

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

Inventory Sales Rate DELIVERY Rate Desired Inventory Inventory Gap Time to Adjust Inventory + Desired Order Rate +

• Expected Sales

Rate Change in Sales Fraction Adjusted + Quantity in Transit Order Rate + Supply Delay +

Inventory model with supply line

• But Supply Line

information is not used in decision making

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

Accounting Supply line in Decisions

• How do we consider the supply line (i.e., past

and pending orders in transit) when we make new ordering decisions?

• Adjust In-transit inventory in reference to a

Desired In-transit

– Similar to how we adjusted inventory levels

• Update Desired Order Rate to include this

adjustment

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

Model with Supply Line Adjusted

• Underlying equations
• Update your model, simulate and observe the results.

– Save model as InventorySL.mdl

Inventory Sales Rate DELIVERY Rate Desired Inventory Time to Adjust Inventory Desired Delivery Expected Sales Rate Change in Sales Fraction Adjusted + Quantity in Transit Order Rate Supply Delay + Adjustment for inventory Adjustment for intransit Desired Intransit quantity Time to adjust intransit +

• Desired Order Rate
• +
• +

+

• +

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

Model with Supply Line Adjusted (2)

• Observe the results
• Does system reach equilibrium?
• What are the equilibrium values of the stocks?

Comment.

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

How much should be in Supply Line

• Typically, how much In-transit do we need to

maintain?

• How to determine the Desired Intransit quantity?
• Desired In-transit quantity

= Desired Delivery*Supply Delay

• Update your model, simulate and observe the

results.

• Save model as Retailer.mdl

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

How much Desired Inventory?

• How can we set the Desired Inventory?
• Typically based on the number of weeks of

inventory coverage desired.

• Note:

– Desired Inventory will be based on the management’s decision – Desired In-transit Quantity to be determined as per Little’s Law

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

Simultaneous Initial Conditions

• The model to start in dynamic equilibrium.
• What should be the initial values of the stocks?

– When initial Sales is 0 kg/day – When initial Sales is 10 kg/day

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

Your Retailer Model to look like this

• In Model >> Settings, set smaller TimeStep until no

significant change in dynamics

– Choose time step to be about one-eighth the value of the smallest time constant in the model

Inventory Sales Rate DELIVERY Rate Desired Inventory Time to Adjust Inventory Desired Delivery Expected Sales Rate Change in Sales Fraction Adjusted + Quantity in Transit Order Rate Supply Delay + Adjustment for inventory Adjustment for intransit Desired Intransit quantity Time to adjust intransit +

• Desired Order Rate
• +
• +

+

• +

+ + Inventory Coverage + +

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

Fulfilling Customer Orders

• The customer orders can be fulfilled only when

we have required inventory. Typically, inventory is a non-negative quantity, representing the physical availability of goods.

– Orders not immediately filled are lost forever.

• Create new variable, Customer Orders (kg/Day)
• Sales Rate = ?
• Anything else to be changed?
• Change in Sales = Customer orders - Expected Sales

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

Complete Retailer Model

Inventory Sales Rate DELIVERY Rate Desired Inventory Time to Adjust Inventory Desired Delivery Expected Sales Rate Change in Sales Fraction Adjusted

• +

Quantity in Transit Order Rate Supply Delay + Adjustment for inventory Adjustment for intransit Desired Intransit quantity Time to adjust intransit +

• Desired Order Rate
• +
• +

+

• +

+ + Inventory Coverage + + Customer Order Rate + Max Sales Rate Min Order Processing Delay

• +

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

Supply chain: Model Distributor

• Till now, we have a model of the Retailer.
• Let’s suppose the Distributor has exactly the

same decision structure as Retailer.

• Replicate Retailer structure for Distributor

– Copy your Retailer.mdl as Retailer-Distributor.mdl – Open Retailer-Distributor.mdl file. – Set Zoom to 50% – Click in the white area, Click Crtl+A and Crtl+C – Click Crtl+V – Press on any black area and Drag the black area to right of the model area. – Let all the variables with ‘0’ in them denote Retailer

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

Link Retailer and Distributor

• The Customer Orders of Distributor is the

Desired Order rate 0 of Retailer

• The Order rate 0 of Retailer is determined by

the Sales Rate of the Distributor

• Simulate and observe the results.

Note: It is good modeling practice to give proper names to variables. Avoid subscripts. Can you think of appropriate names for Distributor echelon?

IEOR, IIT Bombay IE 604: System Dynamics Modelling & Analysis Jayendran Venkateswaran

Customizing Model

• For Retailer:

– Supply Delay = 2 days – Inventory Coverage = 4 days – Time to adjust inventory = 3 days – Time to adjust inTransit = 3 days – Min Order Processing Delay = 1 day

• For Distributor:

– Supply Delay = 4 days – Inventory Coverage = 8 days – Time to adjust inventory = 3 days – Time to adjust inTransit = 3 days – Min Order Processing Delay = 1 day Suppose End Customer Order Rate = 10 Kg/Day, Simulate the model in dynamic equilibrium