Enhanced (R,S,S) policy for periodic review single-item inventory control

• US 5,897,624 A
• Filed: 07/23/1997
• Issued: 04/27/1999
• Est. Priority Date: 07/23/1997
• Status: Expired due to Term

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Ramaswamy, Sanjay Elathur
• Primary Examiner(s): Stamber, Eric W.
• Assistant Examiner(s): Campa, John
• Application Number:
  US08/898,749
• Time in Patent Office: 643 Days
• Field of Search: 364/468.06, 364/468.07, 364/468.13, 364/468.14, 705/28, 705/29, 705/7, 705/8, 705/10
• US Class Current: 705/28
• CPC Class Codes: G06Q 10/087 Inventory or stock management, e.g. order filling, procurement or balancing against orders
• CPC Class Codes: G06Q 10/0875 Itemisation or classification of parts, supplies or services, e.g. bill of materials

Abstract

A computer implemented method provides decision support for how much of an inventory item to order and when to order where the review period is different from a basic time unit for demand realization and updates. For cases where the review period is greater than the basic time unit for describing the system, the (s,S) rule is replaced with an enhanced policy for deciding when to order and how much to order. This policy becomes necessary as in practice items are often reviewed at different frequencies, even though demand realization and inventory updates may be done every day. Given a periodic review inventory system and given a set of (s,S) values, the computer implemented method produces lower average inventory costs for the same achieved service level. Alternatively, for a given average investment in inventory, the method will result in higher service level than the corresponding (s,S) policy.

Claims

1. A computer-implemented method for calculating an order quantity and a delay value for placing an order in an (R,s,S) policy inventory control system, comprising the steps of:

inputting a plurality of forecast data representing a statistical demand for an inventory item over a time duration, and inputting a service level data representing a statistical standard for an inventory position;

calculating a series of first and second reordering points, and a series of order-up-to points for a plurality of time periods within said time duration, based on said input forecast data and service level data;

for each of said plurality of time periods, detecting the inventory position and, if the inventory position is below the first reorder point for the time period then

if the inventory position is below the second reorder point for the time period, computing an order quantity based on said inventory position and the order-up-to-point for the time period, and

if the inventory position is above the second reorder point, computing an order quantity and a delay value, said delay value being computed based on at least said inventory position and said second reorder point,

wherein said order quantity and delay values are used in generating orders for said inventory item.

2. A computer-implemented method for calculating an order quantity and a delay value according to claim 1, wherein the computing of a delay value includes the steps of:

subtracting the second reordering point from the detected inventory position for the time period;

dividing a result of said subtracting by a difference between said first reordering point and said second reordering point; and

multiplying a result of said subtracting and dividing steps by a value representing a time period for reviewing the inventory position.

3. A computer-implemented method for calculating an order quantity and a delay value according to claim 2, wherein the computing of a delay value further includes the step of rounding a result of said multiplying to a predetermined rounding precision.

4. A computer-implemented method for calculating an order quantity and a delay value according to claim 1, wherein the plurality of time periods are indexed by a value t, and wherein the computing of a delay value includes the steps of:

initializing the delay value d_t =0;

setting index i=t;

initializing BY_t =Y_t ;

repeating steps (a) through (c) while (BY_t >

=s1_t)

(a) d_t =d_t +1;

(b) BY_t =BY_t -FM_i ; and

(c) i=i+1,

where BY_t is a computer loop control variable, FM_i is a datum from the plurality of forecast data, and sit is the first reordering point for time period t.

5. A computer-implemented method for calculating an order quantity and a delay value according to claim 1, wherein the plurality of time periods are indexed by a value t, and

wherein the plurality of forecast data representing a statistical demand for an inventory item over a time duration includes a series labeled as FM_t of forecast mean demand data over a lead time plus said time period, a series labeled as FSD_t of forecast standard deviation data over the lead time plus the time period, a series labeled as FML_t of forecast mean demand over the lead time, and a series labeled as FSDL_t of forecast standard deviation over the lead time, and

wherein said step of calculating a series of first and second reordering points and a series of order-up-to points comprises the steps of:

s1_t =FM_t +(k×

FSD_t), and

s2_t =FML_t +(k×

FSDL_t),

where s1_t is the first reordering point and s2_t is the second reordering point for the time period t, and where k is a safety factor based on said input service level.

6. A computer-implemented method for calculating an order quantity and a delay value according to claim 1, wherein said step of calculating a series of first and second reordering points and a series of order-up-to points for a plurality of time periods within said time duration includes inputting a series of predetermined order-up-to points S_t.

7. A computer-implemented method for calculating an order quantity and a delay value according to claim 1, wherein the step of calculating a series of first and second reordering points and a series of order-up-to points for a plurality of time periods within said time duration includes inputting a series of predetermined order quantities O_t, and

wherein said calculating a series of order-up-to points is according to