Chapter 18 Inventory Theory Frederick S. HillierEducation. ∎ Gerald J. Lieberman © 2015 McGraw-Hill All rights reserved. © 2015 McGraw-Hill Education. All rights reserved. Introduction • Scientific inventory management – Mathematical model describes system behavior – Goal: optimal inventory policy with respect to the model – Computerized information processing system maintains inventory level records – Apply the inventory policy to replenish inventory © 2015 McGraw-Hill Education. All rights reserved. 2 Introduction • Demand – Number of units that will need to be withdrawn from inventory • Deterministic inventory model – Used when demand is known • Stochastic inventory model – Used when demand cannot be predicted well © 2015 McGraw-Hill Education. All rights reserved. 3 18.1 Examples • Example 1: manufacturing speakers for TV sets – One speaker needed per TV set • Sets manufactured on continuous production line – Speakers produced in batches • $12,000 setup cost per batch – $10 unit production cost of a single speaker – $0.30 per month holding cost per speaker – $1.10 per month shortage cost © 2015 McGraw-Hill Education. All rights reserved. 4 Examples • Example 2: wholesale bicycle distribution – Distributor purchases a specific bicycle model from the manufacturer and supplies it to various bike shops – Demand is uncertain – Ordering cost includes administrative cost of $2000 and unit cost of $350 per bicycle – $10 per bicycle holding cost – $150 per bicycle shortage cost © 2015 McGraw-Hill Education. All rights reserved. 5 18.2 Components of Inventory Models • Cost of ordering z units – Includes a static cost and a cost per unit • K is the setup cost and c is the unit cost • Holding cost – Represents all costs associated with holding a unit in inventory until it is sold or used • Cost of tied-up capital • Space • Insurance • Protection © 2015 McGraw-Hill Education. All rights reserved. 6 Components of Inventory Models • Shortage cost – Also called unsatisfied demand cost – Cost incurred when demand exceeds available stock • Backlogging: demand not lost but delayed • No backlogging: orders are canceled or met by priority shipment – Revenue may or may not be included in the model © 2015 McGraw-Hill Education. All rights reserved. 7 Components of Inventory Models • Salvage cost – Negative of the salvage value – Included in the holding cost • Discount rate – Accounts for the time value of money • Classification of inventory model based on how often inventory is monitored – Continuous review – Periodic review © 2015 McGraw-Hill Education. All rights reserved. 8 18.3 Deterministic Continuous-Review Models • Economic order quantity (EOQ) model – Stock levels are depleted over time • Replenished by a batch shipment • Basic EOQ model assumptions – Demand rate is constant at d units per unit time – Order quantity Q to replenish inventory levels arrives all at once when inventory drops to 0 – Planned shortages are not allowed © 2015 McGraw-Hill Education. All rights reserved. 9 Deterministic Continuous-Review Models • Reorder point equals demand rate times lead time © 2015 McGraw-Hill Education. All rights reserved. 10 Deterministic Continuous-Review Models • Components of total cost per unit time T – Production or ordering cost per cycle, 𝐾 + 𝑐𝑄 – Holding cost per cycle, ℎ𝑄 2 2𝑑 • Total cost per unit time • Value of Q, Q* that minimizes T © 2015 McGraw-Hill Education. All rights reserved. 11 Deterministic Continuous-Review Models • Cycle time, t* 𝑡 ∗= 2𝐾/𝑑ℎ • For the speaker example: © 2015 McGraw-Hill Education. All rights reserved. 12 Deterministic Continuous-Review Models • The EOQ model with planned shortages – Third assumption of basic EOQ model is replaced: • When a shortage occurs, the affected customers will wait for the product to become available again. Backorders are filled immediately when order quantity arrives to replenish inventory • The EOQ model with quantity discounts – New assumption: • Unit cost now depends on batch quantity © 2015 McGraw-Hill Education. All rights reserved. 13 Deterministic Continuous-Review Models © 2015 McGraw-Hill Education. All rights reserved. 14 Deterministic Continuous-Review Models © 2015 McGraw-Hill Education. All rights reserved. 15 Deterministic Continuous-Review Models • Different types of demand for a product – Independent demand • Bicycle wholesaler experiences this type of demand – Dependent demand • In the TV speaker example: speaker demand varies with TV set demand • Material requirements planning (MRP) – Technique for managing inventory of dependent demand products © 2015 McGraw-Hill Education. All rights reserved. 16 Deterministic Continuous-Review Models • Just-in-time (JIT) inventory management – Emphasizes reducing inventory levels to the bare minimum • Providing items just as they are needed – Focuses on finding ways to reduce setup costs so that order quantities can be small © 2015 McGraw-Hill Education. All rights reserved. 17 18.4 A Deterministic Periodic-Review Model • When demand varies from period to period – EOQ formula no longer ensures a minimum cost solution • Objective: minimize total cost over n periods • Fixed costs are independent of the inventory policy – Minimize total variable costs over the n periods © 2015 McGraw-Hill Education. All rights reserved. 18 A Deterministic Periodic-Review Model • Example given on Pages 815-817 of the text • An algorithm for an optimal inventory policy – An optimal policy produces only when the inventory level is zero © 2015 McGraw-Hill Education. All rights reserved. 19 18.5 Deterministic Multiechelon Inventory Models for Supply Chain Management • Echelon – Each stage in a multi-stage inventory system • Supply chain – Network of facilities that take raw materials and transform them into finished goods at the customer – Includes procurement, manufacturing, and distribution © 2015 McGraw-Hill Education. All rights reserved. 20 Deterministic Multiechelon Inventory Models for Supply Chain Management • A model for a serial, two-echelon system – Seven assumptions in this model outlined on Page 822 of the text • Echelon stock – Stock physically on hand and downstream at subsequent echelons © 2015 McGraw-Hill Education. All rights reserved. 21 © 2015 McGraw-Hill Education. All rights reserved. 22 Deterministic Multiechelon Inventory Models for Supply Chain Management • Optimizing the two installations separately – A flawed approach – Choosing order quantities for installation 2 must account for the resulting costs at installation 1 • Optimizing the two installations simultaneously – Correct approach – Process outlined on Page 826-827 of the text © 2015 McGraw-Hill Education. All rights reserved. 23 Deterministic Multiechelon Inventory Models for Supply Chain Management • Model for a serial multiechelon system – Six assumptions outlined on Page 828 of the text – Difficult to solve for n > 2 – Simplifying approximations normally made to derive a solution • Roundy’s 98 percent approximation property © 2015 McGraw-Hill Education. All rights reserved. 24 © 2015 McGraw-Hill Education. All rights reserved. 25 Deterministic Multiechelon Inventory Models for Supply Chain Management • Extension of serial multiechelon model can be formulated for a distribution system © 2015 McGraw-Hill Education. All rights reserved. 26 18.6 A Stochastic Continuous-Review Model • Traditional method: a two-bin system – All units for a particular product held in two bins – Capacity of one bin equals the reorder point – Units first withdrawn from the other bin – Emptying second bin triggers a new order • Newer approach: computerized inventory systems – Current inventory levels are always on record © 2015 McGraw-Hill Education. All rights reserved. 27 A Stochastic Continuous-Review Model • Inventory system based on: – Reorder point, R – Order quantity, Q • Inventory policy: whenever inventory drops to R units, place an order for Q more units • Ten model assumptions outlined on Page 839 of the text © 2015 McGraw-Hill Education. All rights reserved. 28 A Stochastic Continuous-Review Model • Choosing the order quantity, Q – Use formula for EOQ model with planned shortages • d is the average demand per unit time • See assumptions for definitions of K, h, and p © 2015 McGraw-Hill Education. All rights reserved. 29 A Stochastic Continuous-Review Model • Choosing the reorder point, R – Based on manager’s desired level of service to customers • Alternative measures of service level – Probability that a stockout will not occur between the time an order is placed and when the order quantity is received – Average number of stockouts per year © 2015 McGraw-Hill Education. All rights reserved. 30 A Stochastic Continuous-Review Model • Alternative measures of service level – Average percentage of annual demand that can be filled immediately – Average delay in filling backorders when a stockout occurs – Overall average delay in filling orders • Where delay without a stockout is zero © 2015 McGraw-Hill Education. All rights reserved. 31 A Stochastic Continuous-Review Model • Measure 1 is most convenient to use • Procedure for choosing R under service level measure 1 – Choose L – Solve for R such that 𝑃(𝐷 ≤ 𝑅) = 𝐿 • Example given on Page 841 of the text © 2015 McGraw-Hill Education. All rights reserved. 32 18.7 A Stochastic Single-Period Model for Perishable Products • Stable product – Will remain sellable indefinitely • Perishable product – Can be carried in inventory only a certain amount of time – Single period model is appropriate in this case – Types of perishable products • Newspapers, flowers, seasonal greeting cards, fashion goods, and airline reservations for a particular flight © 2015 McGraw-Hill Education. All rights reserved. 33 A Stochastic Single-Period Model for Perishable Products • Seven assumptions of the model – Given on Pages 846-847 of the text • Analysis of the model with no initial inventory and no setup cost – Simplest case to consider – See Pages 847-849 – Application to the bicycle example • Analysis extends to include setup cost and initial inventory levels © 2015 McGraw-Hill Education. All rights reserved. 34 18.8 Revenue Management • Airlines started using revenue management in the late 1970s • Overbooking – One of the oldest and most successful revenue management practices • Revenue management in the airline industry today – Pervasive, highly developed, and effective © 2015 McGraw-Hill Education. All rights reserved. 35 Revenue Management • Model for capacity-controlled discount fares – Decision variable: inventory level that must be reserved for highest-paying customers – Key to solving: marginal analysis • An overbooking model – Choose overbooking level to maximize profit – Shortage cost (denied-boarding cost) is incurred if overbooking level is too high © 2015 McGraw-Hill Education. All rights reserved. 36 18.9 Conclusions • Models presented in this chapter illustrate the general nature of inventory models • EOQ models have been widely used • Stochastic single-period model is appropriate for perishable products • Multiechelon inventory models play an important role in supply chain management © 2015 McGraw-Hill Education. All rights reserved. 37