2.4 Managing Unpredictable Variability Flashcards
Demand Forecast
Basic assessment of future demand based on
statistical methods
▪ Based on historical numbers/data (on demand)
▪ Statistical or causal models
Demand Plan
Complementing the demand forecast by expert knowledge and confirmation
▪ Based on demand forecast
▪ Complemented by expert knowledge from other corporate divisions
Fig. Production Management
Forecasting in the Supply Chain Context
▪ Coordinated forecast by all members of the supply chain
▪ Use point-of-sales data
Value of Improved Forecasts
▪ Decisions regarding ordering quantities are based on a demand forecast
▪ If demand exceeds supply → Stockouts result from the actual demand (lost sales)
▪ If supply exceeds demand → Overstocks increase the supply chain costs
→ Precise forecasts reduce costs of overstocks and understocks
Forecast Quality
▪ Normally distributed demand:
Mean value µ and standard deviation / forecast error 𝜎
▪ Measures:
Reduction of forecast error, for instance by improving market research leads to higher supply chain profits
→ Reach the same CSL with a lower safety inventory
Conclusion – Characteristics of Forecasts
▪ Forecast are always inaccurate
→ Include both the expected value of the forecast and a measure of forecast error
▪ The farther up the supply chain a company is (distance to the consumer), the greater the distortion of information it receives
▪ Collaborative / coordinated forecasts of demand along the supply chain
▪ Aggregate forecasts are more accurate than disaggregate forecasts
→ Aggregation over regions, products, components, etc.
▪ Accuracy of forecasts decreases with the length of the forecasting period
→ Reducing supply lead time allows using accurate short-term forecasts
Aggregation to Reduce Variability
Aggregating (safety) inventory for various customers / regions in centralized storage facilities in order to take
advantage of compensatory effects.
Comparison of Aggregated and Separate Orders
→ Lower values for the demand correlation 𝝆𝒂𝒃 imply higher potential savings from aggregation of orders.
Impact of Aggregation on Safety Inventory
Square-root Law: Reducing the number of independent stocking locations by a factor of 𝑛, the average safety inventory can be reduced by a factor of sqrt(𝑛).
→ However, response time and transportation cost to customers increase!
→ Alternative types of aggregation, which enable taking advantage of reduced safety inventories without having to physically centralize all inventories in one location.
Types of Aggregation
- Information centralization – aggregation of information
- Specialization – aggregation of inventory
- Product substitution – aggregation of demand
- Component commonality – aggregation of components
- Postponement – aggregation of products
Aggregation of Information
▪ Inventory of every stocking location is globally visible through information systems
▪ Orders are fulfilled by the closest location with stock
▪ In case of delivery problems, another stocking location fulfils the order
▪ Advantages:
– Higher responsiveness
– Lower transportation costs
– Higher product availability
– Lower safety inventory
Aggregation of Inventory
▪ Stocking certain products only in specialized locations
▪ Advisable if:
– Products have different demands depending on the region (e.g., snow shovels)
– Reduction of variety
– Reduction of safety inventory
– Small increase in transportation costs and response time
▪ Products suitable for specialization should have:
– High value
– Low demand
→ Only keep decentralized stock of slow moving products with unpredictable demand if customer is willing to pay a premium!
Aggregation of Demand
▪ Use of one product to satisfy demand for a different product
▪ Manufacturer-Driven Substitution (One-Way-Substitution): Substitution of a lower-value product that is not in stock with a higher-value product
▪ Customer-Driven Substitution (Two-Way-Substitution): Customer substitutes a product that is not in stock with a similar product that is in stock
→ Product substitution enables aggregation and therefore reduces safety inventory
Aggregation of Components
▪ Using a common technological platform for various products
- Advantages:
▪ Lower number of variants for parts → lower development and testing costs
▪ Component Commonality → Economies of scale
▪ Reducing inventory
- Disadvantages:
▪ Greater effort in product design, as common component needs to fit with every product
▪ Changing the common components causes changes in every product
▪ Potentially increased difficulty of product differentiation
▪ Quality issues affect all products which contain common component
Aggregation of Products
▪ Delay product differentiation until closer to the time the product is sold
→ Common components in the push phase, product differentiation as close as possible to the pull phase
▪ Aggregate planning for activities before product differentiation
→ Higher forecast accuracy
▪ Shifting disaggregate forecasts closer to the time of sale
→ More information available
→ Better alignment of supply and demand
Fig. Postponement – Aggregation of Products
Requirements for Postponement
- Modular product design
→ Simple and low-cost assembly of different product variants - Modular manufacturing process
→ Supporting the reorganization of the distribution process - Capabilities of the supply chain
– Efficiently providing the base products
– Flexible and responsive processing of individual customer orders
Enablers of Postponement
- Process Postponement:
→e.g., dye:
Stocking white dye at the retailer, mixing the desired color in the store - Process Resequencing
→ e.g., Benetton:
Dying clothes after the sewing process, that way colors can be changed according to the current trends - Process Standardization
→ e.g., HP drives:
Performing customer specific tests only on receiving a customer order, tests are divided into customer specific and generic (standardized) tests
Potentials and Limitations of Postponement
▪ Increase in profit and better alignment of supply and demand in case of high product variant variability as well as comparable and independent demand for product variants.
▪ If there is a significantly higher demand for a product variant, profits decrease as higher production costs outweigh the advantages of aggregation.
▪ Tailored postponement: Deterministic share of demand is produced cost efficiently without postponement and uncertain share is produced using postponement
Quick response
refers to actions which reduce supply lead time.
▪ Improving Forecast Accuracy with shorter forecasting periods
– Higher number of orders is possible
– Better match supply and demand
– Higher supply chain profitability
▪ However,
– Increased uncertainty for manufacturer, need for flexible production systems
– Manufacturer needs to produce more lots or keep larger inventories
– High costs for setting-up and ordering
