SCM 371 Final Exam (Ch.8)
Lot-sizing with demand that depends on the demand for other items use:
Material Requirement Planning (MRP)
Economic Order Quantity (EOQ)
Q*= Square root (2RS/KC)
EOQ Model 2
R: Demand rate Q: the size of each order T: the time between consecutive ordering opportunities- the quantity we want to find
When do we reorder?
Reorder Point: P=L x (1/250) x R
Who is blamed for overages or storages no matter who made the original forecast or how bad the forecast was?
Supply Managers
Find the optimal interval
T*= Square root (2S/ RKC)
Which of the following statement is false about the inflation of forecasts?
The buyer may discount the supplier's forecast to avoid unnecessarily raising capacity
Fill rate (β or type 2 service level)
The quantity demand that is serviced immediately dividend by the total quantity demanded: β=( total quantity serviced immediately )/ (total quantity demanded )
Service level
There are different metrics for the level or coverage of service. It is important to understand exactly how to organization measures its service level
Customers' demand for cars is an independent demand for a car manufacturer.
True
Exponential smoothing
Use the weighted average of the current demand (yt) and the demand forecast from the previous period y^t to forcast the next period: y^t+1= ayt+ (1-a) y^t (a is the smoothing parameter)
Continuous improvement
We can apply continuous improvement (kaizen) methods to forecasting by tracking forecast accuracy and taking steps to eliminate root causes of forecast error
Time Series
a series of data indexed (or listed or graphed) in time order
When a retailer uses daily sales of each product to identify patterns and to forecast inventory requirements, this is an example of:
a time series forecasting technique
Collaboration between supply chain partners
an efficient method for improving forecasts and service and reducing cost
Enterprise Resource Planning (ERP)
an information system that records various steps in the fulfillment of customer orders, accessible to all areas of the company- manufacturing, finance, sales, marketing, human resources, and supply
Fixed-Period Model: another way to apply EOQ
at times it would be easier to schedule orders at fixed intervals rather than constantly checking whether the stock reaches a reorder point
On an annual requirement of 100 items spread evenly throughout the year, any purchaser has an opportunity of buying all 100 units at a price of $100 each, or buying 10 units at a time at a price of $150 each. The inventory carrying cost is 25 percent per year. There is no ordering costs. Consider the total cost of purchasing and holding the inventory.
buying 100 at a time will save the company $3,937.50 per year: -100 items x 100= 10,000 (RC) -100/2=50 (Q/2) -100 x .25x 50= 1,250 (KC x Q/2) -TC= 11,250 -100 items x 150=15,000 (RC) -10/2= 5 (Q/2) -150x .25x 5= 187.50 -TC= 15,187.50 -15,187.50-11,250= 3,937.50
Expected demand in each period
cycle stock
Fixed-Order-Quantity system
each order is the same fixed quantity
Supply function
often makes forecast
Capacity requirements planning (CRP):
plans for manufacturing resources based on the material plan from MRP
Safety stock (safety inventory, buffer, buffer inventory)
the amount of the target level of inventory in excess of the expected demand in a period: SS=Target inventory- expected demand
Demand for subassemblies, raw materials, and parts will depend on:
the assembly schedule for finished goods
Collaborative Planning, Forecasting, and Replenishment (CPFR)
the business practice of multiple trading partners agreeing to exchange knowledge and share risks to generate the most accurate forecast possible and develop effective replenishment plans: -agree to mutual business objectives and measure -develop joint sales and operations plans -generate and update sales forecasts and replenishment plans
In resale
the buyer may have to asses the expected sales volume (including reduced prices for seasonal goods)
Tradeoff in using safety stock to improve service level
the more safety stock to carry, the higher service level, but also the higher the inventory holding cost
Marketing Responsibility
the need for raw materials, services, and subassemblies usually derives from a sales forecast
α service level (type 1 service level)
the probability that all demand within a period is fulfilled without delay: 1−ℙ{stockout}
Time series forecasting
the use of a model to predict future values based on previously observed values
EOQ cost
-(S x R/Q) + (KC x Q/2) -Sometimes people omit the RC from the objective function since the decision variable Q does not change the total purchasing cost RC
FOQ Costs
-Annual purchase cost (R x C): price paid for the product itself -Annual ordering cost (S x R/Q) -Annual invnetory holding cost (KC x Q/2)
Basic time series patterns
-Constant value (data fluctuate around a constant mean) -Trend (systematic increase or decrease in the mean over time) -Seasonal variations (or cyclical variations) -Turning point (ex. COVID-19) -Random variations
Forecasting Techniques
-Cross-sectional models -Time series models -Wisdom of the crowd
FOQ Model
-Demand rate (R): units per year or per week, assumed constant -Lead time (L): assumed constant -Delivered purchase cost per unit (C): the price the buyer pays the supplier per unit -Order (or setup) cost per order (S): the fix cost of each order -If the buyer orders Q units at a time, it pays S+CxQ for the order -Holding cost rate (K): per dollar per year -When inventory drops to a reorder point (P), a quantity Q is ordered -No back orders or stockouts
Inflation of Forecasts
-Fundamental problem of agency: The supplier bears the cost of the capacity, but the buyer bears the cost of an underage -The buyer has the incentive to inflate the forecast so that the supplier will raise its capacity, reducing the risk of underage -Knowing this, the supplier may discount the buyer's forecast, forcing the buyer to further inflate
Complexity arising from randomness
-Here we only consider the variability in demand. There may be many other significant variabilities that we need to consider for a problem -We resort to a rule of thumb instead of an actual optimal solution. While for a small business this may be fine, a large company in a competitive industry may not afford to stop at merely a rule of thumb
Three key questions
-How much to acquire: quantity -When to acquire?: timing -How to manage inventory effectivity?
Problem 1: Who makes the forecast?
-If sales, production, or user makes a forecast, should supply management second-guess the forecast? -Should other supply chain members be involved in collaborative forecasting?
Types of Demand
-Independent demand: directly from customer demand -Dependent or derived demand: item is part of a larger component or product, and its use is dependent on the production schedule for the larger component
Manufacturing Resource Planning (MRP II)
-Links the MRP and CRP with the financial system -It stress-tests the financial and cash flow strength given possible production scenarios that may arise from the variability in the production process
Push vs. Pull systems
-Push: production systems where companies produce based on a forecast and then "push" the products on the market -Pull: production systems where companies produce based on real-time customer demand
Problem 2: If the forecast is far off, who bears the risk?
-Should suppliers be held responsible for meeting the forecast or the actual requirements? -Should the supply manger be held responsible for meeting the forecast or the actual requirement? -Should the contract assign the responsibility for dealing with inaccurate forecast? -How does negotiation resolve these issues?
Total Cost
-TC= RC+ (S x R/Q) + (KC x Q/2) -The buyer's objective is to minimize the total cost
During a period of ten weeks, a manufacturer received an order for 100 units each week, filled nine of the orders fully and immediately, and filled 60 units for the remaining order immediately (i.e., backlogged 40 units). What is the fill rate during the period?
-Total quantity demanded: 10 x 100= 1,000 -Total quantity served immediately: (90 x 100) +60 -Fill rate= 960 / 1,000= 96%
Wisdom of the crowd
-gather opinions and use with judgement to forecast -the method relies on people with imitate knowledge and a "feel" that is hard to define but may give good forecasting results
Statically significance in a regression model
-shows correlation between the factor and the outcome, may not show causality by itself -when an analyst uses a noncausal model, they must justify the choice of the independent variables
Time series techniques
-simple moving average -weighted moving average -exponential smoothing
Complexity of the model further increases
-when lead times, usable received, inventory shrinkage rates, and so on vary addition to demand -When the distribution of any variable is nonmoral -When the distribution changes with time
The demand for the current period is 69 and the demand forecast made for the current period (made in the previous period) was 67. What is the demand forecast for the next period using exponential smoothing with smoothing parameter 0.3?
.3(69) + (1-.3)67= 67.6
Quantity and timing issues
1. Forecast- Of future demand, lead time, prices, and other costs. Managers must make purchase decisions (long) before requirements are known 2. Costs- Placing orders, holding inventory, stock out, unavailable service 3. Availability- Materials or service may not be available in the desired quantities without paying a higher price or delivery charge 4. Shortages 5. Price-Volume Relationship- Reduced prices for buying larger quantities
Cross-Sectional Models
1. Identify factors that may affect the outcome (demand) to estimate 2. Regress the outcome variable on the factors and controls
Steps of Capacity Requirement Planning (CRP)
1. MRP system develops a material plan 2. CRP translates the plan into the required human machine resources by workstation and time bucket 3. CRP compares the required resources against available resources 4. If capacity is insufficient, adjust either the capacity or the master production schedule- "closed-loop MRP"(while the MRP drives the CRP, the CRP may come back to adjust the marital plan)
Data for Material Requirement Planning
1. Master production schedule: how many end are to be produced during a specified time period 2. Structured bill of materials (BOM): the subcomponents necessary to manufacture one finished item 3. Inventory record: open orders, lead times, and lot-size policy to calculate the quantity and timing of orders
Eliminating poor predictors from the crowd
1. Measure each judge's performance relative to the group 2. Use positive contributors to build a weighting model for aggregating forecasts
Four scenarios of service coverage
1. Only some of the buffer inventory was used 2. No buffer inventory remained but there was no stock out (for continuous distributions of demand almost never the case) 3. There was a stock out. (stock out cost would be incurred) 4. All the buffer inventory remained
Time-based strategies
1. Reduce setup and cycle times -Long lead times and high setup costs may force the buyer to buy large quantities and keep inventory high 2. Coordinate the flow of resources -Eliminate process/ system waste -Just-on-time or JIT
Statistical significance in a regression model proves a cause-and-effect relationship between a factor and the quantity of sales
False