JUST-IN-TIME (JIT) PRODUCTION

Just-in-time (JIT) is defined in the APICS dictionary as “a philosophy of manufacturing based on planned elimination of all waste and on continuous improvement of productivity”.  It also has been described as an approach with the objective of producing the right part in the right place at the right time (in other words, “just in time”).  Waste results from any activity that adds cost without adding value, such as the unnecessary moving of materials, the accumulation of excess inventory, or the use of faulty production methods that create products requiring subsequent rework.  JIT (also known as lean production or stockless production) should improve profits and return on investment by reducing inventory levels (increasing the inventory turnover rate), reducing variability, improving product quality, reducing production and delivery lead times, and reducing other costs (such as those associated with machine setup and equipment breakdown).  In a JIT system, underutilized (excess) capacity is used instead of buffer inventories to hedge against problems that may arise.

JIT applies primarily to repetitive manufacturing processes in which the same products and components are produced over and over again.  The general idea is to establish flow processes (even when the facility uses a jobbing or batch process layout) by linking work centers so that there is an even, balanced flow of materials throughout the entire production process, similar to that found in an assembly line.  To accomplish this, an attempt is made to reach the goals of driving all inventory buffers toward zero and achieving the ideal lot size of one unit.

The basic elements of JIT were developed by Toyota in the 1950's, and became known as the Toyota Production System (TPS).  JIT was well-established in many Japanese factories by the early 1970's.  JIT began to be adopted in the U.S. in the 1980's (General Electric was an early adopter), and the JIT/lean concepts are now widely accepted and used.

Some Key Elements of JIT

1. Stabilize and level the MPS with uniform plant loading (heijunka in Japanese): create a uniform load on all work centers through constant daily production (establish freeze windows to prevent changes in the production plan for some period of time) and mixed model assembly (produce roughly the same mix of products each day, using a repeating sequence if several products are produced on the same line).  Meet demand fluctuations through end‑item inventory rather than through fluctuations in production level.  Use of a stable production schedule also permits the use of backflushing to manage inventory: an end item’s bill of materials is periodically exploded to calculate the usage quantities of the various components that were used to make the item, eliminating the need to collect detailed usage information on the shop floor.

2. Reduce or eliminate setup times: aim for single digit setup times (less than 10 minutes) or "one‑touch" setup ‑‑ this can be done through better planning, process redesign, and product redesign.  A good example of the potential for improved setup times can be found in auto racing, where a NASCAR pit crew can change all four tires and put gas in the tank in under 20 seconds.  (How long would it take you to change just one tire on your car?)  The pit crew’s efficiency is the result of a team effort using specialized equipment and a coordinated, well-rehearsed process.

3. Reduce lot sizes (manufacturing and purchase): reducing setup times allows economical production of smaller lots; close cooperation with suppliers is necessary to achieve reductions in order lot sizes for purchased items, since this will require more frequent deliveries.

4. Reduce lead times (production and delivery): production lead times can be reduced by moving work stations closer together, applying group technology and cellular manufacturing concepts, reducing queue length (reducing the number of jobs waiting to be processed at a given machine), and improving the coordination and cooperation between successive processes; delivery lead times can be reduced through close cooperation with suppliers, possibly by inducing suppliers to locate closer to the factory.

5. Preventive maintenance: use machine and worker idle time to maintain equipment and prevent breakdowns.

6. Flexible work force: workers should be trained to operate several machines, to perform maintenance tasks, and to perform quality inspections.  In general, JIT requires teams of competent, empowered employees who have more responsibility for their own work.  The Toyota Production System concept of “respect for people” contributes to a good relationship between workers and management.

7. Require supplier quality assurance and implement a zero defects quality program: errors leading to defective items must be eliminated, since there are no buffers of excess parts.  A quality at the source (jidoka) program must be implemented to give workers the personal responsibility for the quality of the work they do, and the authority to stop production when something goes wrong.  Techniques such as "JIT lights" (to indicate line slowdowns or stoppages) and "tally boards" (to record and analyze causes of production stoppages and slowdowns to facilitate correcting them later) may be used.

8. Small‑lot (single unit) conveyance: use a control system such as a kanban (card) system (or other signaling system) to convey parts between work stations in small quantities (ideally, one unit at a time).  In its largest sense, JIT is not the same thing as a kanban system, and a kanban system is not required to implement JIT (some companies have instituted a JIT program along with a MRP system), although JIT is required to implement a kanban system and the two concepts are frequently equated with one another.

Kanban Production Control System

A kanban or “pull” production control system uses simple, visual signals to control the movement of materials between work centers as well as the production of new materials to replenish those sent downstream to the next work center.  Originally, the name kanban (translated as “signboard” or “visible record”) referred to a Japanese shop sign that communicated the type of product sold at the shop through the visual image on the sign (for example, using circles of various colors to indicate a shop that sells paint).  As implemented in the Toyota Production System, a kanban is a card that is attached to a storage and transport container.  It identifies the part number and container capacity, along with other information, and is used to provide an easily understood, visual signal that a specific activity is required.

In Toyota’s dual-card kanban system, there are two main types of kanban:

1. Production Kanban: signals the need to produce more parts

2. Withdrawal Kanban (also called a "move" or a "conveyance” kanban): signals the need to withdraw parts from one work center and deliver them to the next work center.

In some pull systems, other signaling approaches are used in place of kanban cards.  For example, an empty container alone (with appropriate identification on the container) could serve as a signal for replenishment.  Similarly, a labeled, pallet-sized square painted on the shop floor, if uncovered and visible, could indicate the need to go get another pallet of materials from its point of production and move it on top of the empty square at its point of use.

A kanban system is referred to as a pull‑system, because the kanban is used to pull parts to the next production stage only when they are needed.  In contrast, an MRP system (or any schedule‑based system) is a push system, in which a detailed production schedule for each part is used to push parts to the next production stage when scheduled.  Thus, in a pull system, material movement occurs only when the work station needing more material asks for it to be sent, while in a push system the station producing the material initiates its movement to the receiving station, assuming that it is needed because it was scheduled for production.  The weakness of a push system (MRP) is that customer demand must be forecast and production lead times must be estimated.  Bad guesses (forecasts or estimates) result in excess inventory and the longer the lead time, the more room for error.  The weakness of a pull system (kanban) is that following the JIT production philosophy is essential, especially concerning the elements of short setup times and small lot sizes, because each station in the process must be able to respond quickly to requests for more materials.

Dual-card Kanban Rules:

  1. No parts are made unless there is a production kanban to authorize production.  If no production kanban are in the “in box” at a work center, the process remains idle, and workers perform other assigned activities.  This rule enforces the “pull” nature of the process control.
  2. There is exactly one kanban per container.
  3. Containers for each specific part are standardized, and they are always filled with the same (ideally, small) quantity.  (Think of an egg carton, always filled with exactly one dozen eggs.)

Decisions regarding the number of kanban (and containers) at each stage of the process are carefully considered, because this number sets an upper bound on the work-in-process inventory at that stage. For example, if 10 containers holding 12 units each are used to move materials between two work centers, the maximum inventory possible is 120 units, occurring only when all 10 containers are full.  At this point, all kanban will be attached to full containers, so no additional units will be produced (because there are no unattached production kanban to authorize production).  This feature of a dual-card kanban system enables systematic productivity improvement to take place.  By deliberately removing one or more kanban (and containers) from the system, a manager will also reduce the maximum level of work-in-process (buffer) inventory.  This reduction can be done until a shortage of materials occurs.  This shortage is an indication of problems (accidents, machine breakdowns, production delays, defective products) that were previously hidden by excessive inventory.  Once the problem is observed and a solution is identified, corrective action is taken so that the system can function at the lower level of buffer inventory.  This simple, systematic method of inventory reduction is a key benefit of a dual card kanban system. 


These lecture notes are intended to supplement the assigned textbook for this course (currently, Chase, Jacobs, and Aquilano, 11th edition, Irwin/McGraw-Hill, referred to as "CJA") and often are based very closely on the associated chapter(s) in the textbook.   Significant portions of these lecture notes were originally based on two early articles by Richard J. Schonberger: "Just-In-Time Production Systems: Replacing Complexity With Simplicity in Manufacturing Management", Industrial Engineering, October 1984, pages 52-63; and Applications of Single-Card and Dual-Card Kanban, Interfaces, August 1983, pages 56-67.  This page was last updated on January 27, 2006.