TOC and MRPII

TOC and MRPII, From Theory to Results
Theory of Constraints and MRP II: From Theory to Results
Jeffry J. Smith
smithjj@cat.com
May 11, 1994
Bradley University, Peoria Illinois
ABSTRACT
This paper presents the use of Theory of Constraints (TOC) in
conjunction with Material Requirements Planning (MRP II) for shop floor
scheduling. The differences between TOC, Just-In-Time (JIT) and MRP are
defined. A hybrid MRP and TOC system is presented. Then the
implementation of TOC is covered, from both theoretical and empiric
perspectives. Finally, production results from several implementations
of TOC are documented. The conclusion of this paper is that TOC offers
manufacturers in job shop and repetitive environments a significant
improvement in shop floor productivity, on time delivery, and
predictability. However, the achievement of such gains requires
considerable preliminary training, and policy changes, first in
management, then on the shop floor.
INTRODUCTION
MRP II is a planning system, with the goal of delivering the correct
quantity of material at the correct time, based upon orders combined
with forecasts. MRP II has been widely used in the U.S. with many
cases of success and improvement. (Vollmann)
TOC is a management philosophy, with the goal of making money, in both
the present and future. This goal is accomplished by 1) maximizing
throughput (rate of sales); 2) by minimizing inventory. (material cost
of goods not yet sold); and 3) by minimizing operating expenses, all
money spent to turn inventory into throughput. This includes direct
labor, overhead, and all capital goods. Operationally, maximizing
throughput means ensuring the market, not the factory, is the
constraint of additional sales.
A key insight of TOC is that only a few work centers within the factory
control the output of the entire factory for each product line.
Managing these capacity constraining resources (CCRs) or bottlenecks
optimizes the output of the factory. Knowledge of the plant's CCRs also
provides guidance for future plant investment.
TOC first became known in the U.S. through a shop floor planning and
control program known as OPT, sold by Creative Output beginning in
1979. (Jacobs). The program was developed by Eli Goldratt, who later
expanded its principles into the Theory of Constraints, as explained in
his books, The Goal, The Race, The Haystack Syndrome, and The Theory of
Constraints. The principles embodied in OPT are completely subsumed

by TOC.
Because MRP II arrived before TOC, TOC has generally been installed
with an MRP II system already in place. Reviewing the literature, TOC
is generally viewed as complementary to MRP, supplying finite shop
floor scheduling while MRP generates the overall demand. (Vollmann,
Ptak, et.al). Vollmann views the scheduling program OPT as an
enhancement to MRP's management production control system (MPC).
There have been some spectacular successes with TOC, and some
abandonment of the technique. This paper focuses on the distinctive
aspects of TOC compared with MRP and JIT and how to implement it
successfully under an MRP II system.
THEORY OF CONSTRAINTS: KEY DIFFERENCES FROM MRP II
Some of TOC advantages are that no large scale equipment movement or
computer system changes are needed. The big change with TOC is in
focus. All accounting burden is put on the CCR resources. The CCR
utilization is used to determine which products maximize throughput. A
"good" product is one that has a high contribution and uses little of
the CCR resource. (Bakke)
Page 1
TOC builds on MRP II data requirements, but does not require extreme
data accuracy except at points feeding the CCR resource. (Ptak, Ronen &
Starr) The Pareto rule used to determine CCRs, and consequently which
data needs to be accurate for decision making. The CCR and its feeder
data must be accurate, or OPT/TOC won't work. (Jacobs) The
identification of CCRs also shows where setup reductions and process
improvement efforts should be focused. This is part of the 9 OPT
rules. (See Appendix) (Ronen & Starr).
OPT's scheduler allows for more constraints than MRP. This permits the
MRP and Capacity Resource Planning (CRP) functions to be merged into one
production planning tool. (Plenert) OPT includes rough cut capacity
planning, which when combined with the machine resource data can
estimate the capacity required at each work center. The average loads
on machine centers are sorted in descending order and the most heavily
loaded are studied. (Vollmann)
MRP is a infinite scheduler. OPT creates finite schedules for only the
CCR operations. The downstream operations are finite forward loaded
based upon the capacity of the CCR resource. The upstream operations
are back scheduled from the CCR using MRP logic. (Vollmann) This
procedure greatly simplifies finite scheduling. (Spencer) The
resulting schedules can be modified quickly in a few hours. The net
effect of the finite scheduling and a key contribution of OPT is an
achievable master production schedule (MPS). By finite scheduling only
a few work centers, there are few priority conflict issues between the
MRP schedule and the finite loading.
Whereas MRP assumes that setup, move, and processing times are
deterministic, TOC assumes nothing is deterministic. A unique and key
insight of TOC is that TOC approaches planning like PERT, trying
to find the critical path in the factory routing and manage the

critical resources. Any rules based on "average" times fail because of
statistical fluctuations. In MRP dependent events combined with
statistical fluctuations natural to the factory environment reduce
throughput and increase WIP. This produces "waves" of inventory in a
factory, and the month end "rush" symptoms. (Koziol)
Lead times are the result of a TOC schedule and cannot be pre
determined. In practice, this means lead times are variable, unlike
MRP. As a start-up heuristic, TOC sets lead times to be three times
processing time. This is one third of most MRP systems' lead times, and
is due to smaller batch sizes, less WIP, and less queue time. TOC also
uses operation overlapping as much as possible. Finally, TOC expedites
material not available when two thirds of the lead time has transpired.
Additionally, with OPT/TOC the transfer batch should be optimized to
maximize throughput, not automatically set to the process batch size,
as is done in MRP. The process batch itself should be variable, a
function of the schedule, potentially varying by operation and over
time. Both lot sizes vary, with the goal of maximizing production
across the CCR.
OPT is designed to split orders on the shop floor. However, rather
than splitting the orders on the CCR or back logged machine, as is
normally done, the orders are split on the non- CCR machines, where
more setups are done. OPT creates larger lot sizes on the CCR machines
and smaller lot sizes on non-CCR machines. The transfer batches are
usually smaller than the process batches, since OPT uses overlapping
schedules to maximize production and reduce lead time.
Page 2
An interesting illustration of the difference between OPT and MRP is
the case when there are no factory constraints, that is, if the market
is the constraint. Like MRP, OPT will back schedule the factory from
the market demand. However, OPT will reduce the batch sizes until some
resources almost become CCRs. This reduces WIP and lead time and
increases material velocity.
The production schedule is driven not by orders, nor by the MPS, but by
the "Drum", the pace of the factory CCR. When the market is the
constraint, orders act as the drum, and material is pulled through the
factory.
A time buffer and safety stock are established in front of this CCR and
any critical operations that feed the CCR. This is to protect the CCR
from statistical fluctuations which would stop this critical process.
Another buffer is established before the final assembly line. A safety
stock of CCR parts is kept here. Shipping is also buffered from
variability in the assembly line.
Plant work centers other than constraints are not scheduled. The work
rule may be: "work if you have materials; do not work if you do not
have materials". (Gardiner) These non-constraining resources have no
inventory buffers. These resources protect against shortages through
excess or safety capacity, since, by definition, these are not CCRs.
This selective buffering of TOC is in contrast to JIT, which puts

inventory (in kanbans) throughout the plant, and MRP, which puts
inventory throughout the plant by launching orders in excess of the
factory's CCR capacity.
A non-CCR resource is made up of three time elements: run time, setup
time, and idle time. Saving time at non-CCR resource only increases
its idle time, and does not increase throughput, reduce inventory, or
operating expense. Unlike JIT, TOC does not seek setup improvement
everywhere. In contrast to MRP, TOC does seek to change setup times
when throughput can be improved.
Unlike standard cost accounting, factory effectiveness under TOC is not
measured by machine efficiency. Efficiency or utilization is made up
of two components: breakdowns or absenteeism and lack of work. The
ideal utilization of the first is 100%. The ideal utilization of the
second is 0%. (Black)
Like JIT, TOC assumes a stable environment. When the factory is
unstable, non-CCRs can become CCRs. In such a variable environment,
one begins by saving time on CCR resources, and then proceeding to
resources with next highest utilization, for it may become the next
CCR. TOC is actually less sensitive to changes in the production plan
than JIT. Production plan changes are analyzed in terms of their
impact upon the CCR, which highlights any problems. Simulation of the
factory can also be carried out in the scheduling process. (Plenert &
Best) Capacity changes in the CCRs can also be simulated. This permits
the planner to anticipate impact of changes in demand, schedules or
work center capacities. JIT, on the other hand, reacts to the actual
results of the change. (Lambrecht)
The time offset between scheduling the CCR and the release of raw
materials is known as the "rope". This time is computed by adding the
processing times of all the resources upstream from the CCR. The
combination of the Drum-Buffer-Rope" constitutes TOC's shop floor
Page 3
scheduling. This is in contrast with JIT which pulls material through
physical signals, and MRP, which releases material constrained only by
the rough cut capacity planning system.
TOC is a push system downstream from the CCR and a pull system upstream
from the CCR. Obviously, if the market is the CCR, then the whole
factory is a pull system, as it is for JIT. But TOC is flexible, and
the CCR may be located anywhere in the factory. MRP treats all
resources as infinite in capacity, and follows the drum of orders only,
pushing material through the factory.
TOC/OPT can create schedules down to hours and minutes of a day at the
CCR's, in contrast with MRP, which works with weekly, or at best, daily
buckets. (Jacobs)
A COMBINED SYSTEM OF MRP-OPT
A combined system of MRP-OPT uses MRP to schedule and order materials.
The planning system is a modified MRP system, with an aggregate
production plan, a stable MPS, and a bill of materials for each

product, and a set of time-phased shop orders. The capacity
requirements plan and the rough cut capacity are fed from the capacity
analysis done from the shop floor through OPT. The MPS is driven not
by customer orders, but by the CCR.
The execution system is an OPT or OPT like system with three levels of
scheduling: 1) master scheduling, 2) material requirements planning,
and 3) shop floor control, which is done through the drum-buffer-rope
(DBR) system.
Planning Phase in a Hybrid System
The lot size is determined by the OPT and fed to MRP. The level mix of
products is determined by the throughput per hour of each product at
the CCR. The CCR cycle time determines the pace of production for the
whole factory. The Master Scheduler uses this pace as the controlling
element in the creation of the master production schedule (MPS).
Material Requirements Planning
MRP receives the customer orders and explodes requirements, but doesn't
push low level orders to the shop floor. MRP only produces a daily
schedule for the final assembly cells, based upon the pace of the
constraining resource.
The CCR "pulls" material release into the plant, based upon the CCR's
pace. The material is pulled to the CCR. Material availability
checking and shop papers are not needed.
Inventory levels are adjusted based upon material deliveries and
completed, shipped production. Material is tracked as it enters and
leaves the plant, the "four walls" inventory system (Schoenberger).
MRP explodes the BOM for the items shipped and backflushes the
component inventory maintained from material delivery receipts.
Internal tracking is eliminated because of the reduction of WIP, and
the consequential acceleration of product through the plant.
Page 4
Capacity Requirements Planning
The entire factory can be treated as a single work center. The factory
loading is determined by the CCR process. There is room for error in
capacity in hybrid system, since the CCR controls the release material
into the factory. However, if the CCR has greater capacity than
planned, another work center may really be the CCR, and the pace or
capacity of the factory will be determined by a different CCR.
Buffer Stocks
One must buffer shipping from final assembly. This time buffer is to
protect shipping from variability on the assembly line. Assembly must
also be buffered from manufacturing. This prevents any interruptions
of the final assembly schedule through variations in the manufacturing
process or in variability from the suppliers. In both these cases,
care must be exercised to add the buffer queue time to only those parts
that enter the final assembly area from the manufacturing area rather

than all parts throughout the plant. (Spencer)
Finally, a buffer must be created at the CCR operation. Like the other
buffers, this is a time buffer. Note that most of these buffers
already exist within the planning phase of MRP. Now their value is
recognized, and these buffers will be actively managed to reduce buffer
size as process improvement efforts in the factory are effected. All
other buffers in the plant can be removed, resulting in a reduction of
overall plant inventory, a decrease in queue time, and a reduction in
lead time.
Execution Phase in the Combined System
The finite schedule at the CCR must be strictly followed. The shop
schedule at the CCR is managed by careful monitoring of the actual
buffer versus the planned buffer. The MP scheduler plans material, but
the plant produces based upon the pace of the CCR area.
The shop floor and vendors produce and deliver exactly the need of
components or parts at the CCR resource. Downstream operations
immediately complete operations as the material moves to them.
The final assembly schedule is prepared to meet the actual ship dates.
Purchasing continues to execute purchase orders that were developed in
the MRP system. The signal for material replenishment is consumption at
the CCR, not anticipated demand.
Shop floor Control
Finite scheduling is simplified through TOC by scheduling only the CCR
operations and deriving upstream and downstream schedules from these.
(Spencer) The goal is not to schedule each machine center but to
execute the finite schedule at the CCR.
Materials from the gating operation are pulled to the CCR. This is done
by communicating the finite production schedule at the CCR back to the
first operation supporting that CCR. The communication link is formed
by eliminating all lead time from the CCR to the first operation in the
MRP system. With zero lead time, the first operation is always past
due. (Spencer)
The goal is to balance flow, not capacity. The factory pace is driven
by the CCR. The completed CCR material is pushed through the
downstream operations and out to the shipping dock.
Page 5
Minimized Set-up Times
Since the output of the whole factory hinges on the productivity of the
CCR, set-up times on the CCR must be driven to absolute minimums.
Since non-CCR work centers no longer need to be run at high
utilization, personnel can be brought from these areas to assist in
set-up time reduction.
Purchasing

Purchasing lead times are offset from the CCR so that variability from
the supplier is buffered. The quantity and timing of deliveries are
based upon MRP explosion. The trigger for shipment is the pull signal
from the CCR.
The key element to watch with vendors is that the vendor lead time does
not become a constraint. Under TOC the entire factory will become more
efficient until the market, vendor, or other external agent is the
constraint upon the factory's productivity. With this in mind, the
company may wish to establish a long term relationship with its vendors
and aid them in applying TOC.
Inventory Control
Overall inventory is reduced through reduction of variability and
uncertainty. Buffers are typically divided into 3 parts. Region 3 is
where any need is ignored. This region may have zero inventory.
Region 2 requires that one locates any missing parts. The parts are
not yet expedited, but are tracked, and action is planned. Region 1
are missing parts near the time of consumption, threatening output of
the plant. These are expedited. The size of each region depends on
the speed of reaction of the plant and the expediter. (Schragenheim &
Ronen)
Quality Improvement
The identification of CCR resources immediately signals these areas as
candidates for setup time reduction. The management of the CCR,
assembly, and shipping buffers tracks the reasons for expediting. The
main causes of shortages are identified as areas for improvement.
These may be Capacity Constrained Resources, work centers that are not
CCRs, but can become one under some conditions.
As the quality improvement process reduces variability, the buffer size
can shrink. The heuristic used is that 90% of all parts are completed
with no expediting. More frequent expediting indicates a need for a
larger buffer. (Gardiner)
Summary
1. The combined MRP-TOC system incorporates finite shop floor
scheduling and capacity-planning aspects in MRP.
2. The combined system avoids the problem of "waves" of inventory and
the "Month end syndrome"
3. The combined system provides a detailed shop floor schedule at the
critical control points. The schedule can be accurate down to minutes.
4. The combined system does not require much additional programming or
any additional capital equipment.
Page 6
IMPLEMENTATION

The key to TOC implementation under MRP is management education.
Goldratt's discussion on the organizational education requirements,
taken from his book Theory of Constraints is summarized below. Spencer
has well defined the details of implementing MRP and OPT. His work is
combined with Goldratt's material. Other authors' comments on
implementation are added in as appropriate.
1. Top Down Education and Implementation
With TOC there is a need for education throughout the company and a
change in overall management philosophy. Who should be trained? The
person with profit and loss responsibility for a functional unit of the
company and the comptroller. These two then plan the sequence of
implementation based upon who in the plant can elevate throughput the
most. The implementation plan is for the next two levels management.
The lower levels plan the next levels of implementation. This process
continues down the organizational structure until a level is reached
where people no longer resent being lead. If people don't want to be
lead, then they must be trained in TOC. (Goldratt)
All members of the management team for all functional areas must buy
into TOC. The biggest potential gain from TOC is increased throughput,
and any functional area that does not adopt TOC will become the
bottleneck for the whole organization. This area will be overworked
and the other functional areas will have excess capacity.
Each step of implementation must not cause any resistance in other
groups, resistance that will jeopardize the implementation of the next
steps. The emotional resistance to change can be conquered by the
stronger emotion of idea ownership. One creates ownership of an idea
by inducing people to invent the idea themselves, through the Socratic
method, ie, using leading questions. (Goldratt)
Bottom Up Implementation
If lower level personnel wish to implement TOC, they need to have two
or more people from different functional areas read the Goal and be
trained in TOC. Then they write a joint enthusiastic letter to a
person with profit and loss responsibility, encouraging him or her to
look into TOC. A general business manager will realize the
significance of two people from different functional areas recommending
the same philosophy and approach and will be likely to investigate.
Once this person is "sold" on TOC and trained in it, implementation
proceeds as in the top down implementation. (Goldratt)
2. Change in Measurements
Measurements are no longer based on machine or man utilization, but on
material flow, specifically the material that can be sold, throughput.
The best way to maximize labor assets is cross training so labor may
move from work center to work center. OPT links clearly its philosophy
on cost accounting and shop floor scheduling and practices. The change
in manufacturing practices is sometimes counter-intuitive and difficult
to implement without support in TOC from top management at the plant.
To get shop floor results congruent with the goal of maximizing

throughput, Goldratt has proposed two new measures to replace
utilization and efficiency: 1) Inventory dollar-days measures the
extent to which a department or worker contributes to an early finish
of an order. This is the number of days of early release of material
into the system, times the dollar value of the inventory.
Page 7

2) Throughput dollar-days measures the lateness of an order. It is the
dollar value of the order times the days lateness. This measurement
also affects quality. Any quality defect that requires rework would be
charged throughput dollar- days. The effect of these measurements is to
effect on time release and completion of work at all work centers,
while reducing WIP. (Gardiner)
OPT implementation requires perfect data accuracy at CCR operations,
and less accuracy at non-CCR areas. Since this is less data accuracy
than what is needed for the existing MRP system, most implementations
will be satisfied by double checking data for the CCRs.
OPT requires changes in paperwork flow to allow easy order and lot
splitting and re-combining.
3. Identify bottleneck operations in the shop
The can be done through MRP reports on work force and machine loads, or
by the simply walking through the shop and observing where inventory
has piled up. Note that MRP often under- reports loads because setup
time is not considered in the load. The shop floor personnel and Master
Scheduler are also good sources for insight on the shop's bottlenecks.
There may also be a list of machines that are candidates for capacity
improvement. The initial set of bottlenecks will change as TOC is
implemented, so perfect accuracy is not necessary.
The pareto rule is used to determine CCRs, and explore candidates for
setup improvement.(Bonan & Starr) The 20% most heavily loaded work
areas are assessed as CCRs.
Generally companies don't know their capacity or CCRs. The
possibilities of extending CCR capacity within short time horizons
through overtime or outsourcing should be taken into account when
identifying the CCR. (Nils, Bakke, Hellberg)
TOC implicitly assumes a stable system in determining CCRs. A plant
should have a stable order mix and given resources before implementing.
(Ronen & Starr)
4. The MPS is Controlled by the CCR
The pace of the factory will be set by the cycle time of the CCR. The
Master Scheduler will accept no more requirements beyond that which can
be produced at the bottleneck during a time period. The master
schedule for the CCR is used to derive the schedule for the MRP system.
5. Developing the Master Schedule
The Master Scheduler develops a finite schedule for the CCR. This can
be done through the OPT software package, another finite scheduling
package, or manually. A simple technique is to start from the shop
flow orders for the CCR as prepared originally by the MRP system.
Past due and current orders will likely cause an overload at the CCR.
This overload can be handled through expediting, overtime, or working
down existing WIP and finished goods inventory.

Regardless of the technique, an achievable plan needs to be made. Once
the CCR plan exists, preceding operations can be back scheduled to
deliver the proper parts to the CCR at the proper times.
Page 8
6. Creating Buffers
First create a buffer before shipping area, to protect the shipping
schedule from variability in the final assembly area. In most MRP
systems, this buffer already exists and can be used as is. Through
buffer management, this buffer and the others will be continually
reduced in size as variability is reduced.
Next, create a buffer before the assembly area, to protect the assembly
line from variability in delivery purchased and manufactured components
needed by the line. This buffer is usually already established through
a lead time parameter in MRP. This lead time buffer should only
contain items that are needed for final assembly, not for all parts
throughout the plant. The routing file should be reviewed to
determine which parts need to be buffered.
Finally, a time buffer is created before the CCR operation. This is
done by adjusting the lead time in the MRP part master for the parts
routed across the CCR.
In all cases, the initial buffer size can be determined by the
heuristic that the buffer should be three times the total cycle time,
including setup time. Very often the existing WIP buffer will be
larger than this, and the MRP lead time will have to be reduced.
Through buffer management, these buffers are adjusted in size until 90%
of the parts can be processed without expediting.
7. Creating the "Rope"
The rope assures that the release of material into the plant is
sufficient to support the bottleneck operation. The finite production
schedule at the bottleneck is communicated to the first operation by
eliminating all lead time from the CCR to the first operation. The
first operation will always appear past due. The elimination of lead
time reduces the WIP in an MRP system. Realize there will be no
schedules for non-CCR operations.
8. Execution
The CCR schedule must be rigidly followed. The CCR schedule is
controlled through the buffer management system discussed earlier.
The final assembly schedule is created to support the actual ship
dates. Purchasing continues to execute purchase orders developed in
the MRP system.
The software system changes are the easy part of the implementation.
The critical part is the need for complete understanding among all
members of the factory management team.

RESULTS
Positive results from TOC can come quickly, within months of beginning
the implementation of TOC. Factory-wide implementation can be complete
within a year or two. (Frizelle)
Three examples are given for results from TOC implementation with MRP.
In the first, MK Electric, the OPT program is used to improve customer
responsiveness. In the second, Valmont/ALS, the internal constraints
Page 9
were linked to the MRP system through internally developed software
during a period of declining sales. In the third example, TOC is
combined with MRP and JIT to reduce lead times during a period of rapid
growth.
Company: MK Electric
Problems: MK Electric had a lack of communication between marketing and
manufacturing, poor customer service, and an extremely broad product
line, with over 4,000 products. Only 1/3 of top product lines were
within the target schedule of ten days.
Result: OPT recalculates production schedule in 1.5 hours. WIP was cut
by 80%. On time deliveries went from 30% to 80%. Orders more than 10
days old were down 83%, a reduction of more than a million dollars.
WIP was reduced 33%, by $10 million. Payback was achieved in 12 months
from inventory reduction alone. (Black)
Company: Valmont/ALS
Problems: Custom orders, little standardization of products, low-tech
manufacturing, capital intensive manufacturing. Valmont experienced
"End of Month Syndrome"--they started each month fine, but then had to
work overtime to meet shipments.
They had difficulty forecasting earnings from production control
schedules. Inventory levels were increasing. Obsolete inventory
write-offs, poor inventory turns. 1986 oil bust hurt 1987 orders. MRP
system filled weekly buckets, based on predetermined lead times from
industry standards, using historic data. MRP forced wip through shop
in "waves", exceeding the capacity of the shop. The waves peaked at
the end of the month and drowned the assembly, finishing, and shipping
resources.
Obstacles: Marketing and engineering had to be educated to new shop
scheduling method. Theory had to be applied to real world. Personal
beliefs in cost accounting had to be abandoned.
Implementation
1. Management Education
1.1 Division president initiated TOC education by distributing The
Goal.

1.2. General Manager gave copies to the staff. Read the Race.
1.3. Management discussed the books and how to apply locally.
1.4. Management studied APICS articles about OPT.
1.5. Controller, Operation Manager, and scheduler attended seminars on
"Logistics of Constraint and Finance", engineering and quality.
2. Constraints Identified
Management identified market as an external constraint, with internal
excess capacity. Engineering function was also an external constraint.
Weld assembly areas internally were constraints, with other floating
CCRs.
Page 10
3. Change in Manufacturing Strategy
Management made the decision to stop producing to stock and only
produce to order. This required the factory to build to order in less
than industry standard lead time.
4. TOC Implementation with MRP
4.1. Reduced small parts batches by 40%--flow improved, expediting and
overall chaos diminished.
4.2. The scheduling system would allow for buffers in front of factory
constraints. The size of the buffers would be determined by down time
of non-constrained resources that feed the constraint. A three day
buffer was determined.
4.3. The constraint became the focal point of the organization's
attention. Program, policy, and asset additions were evaluated in
light of the effect upon the constraint. Preventative maintenance and
quality efforts would be intensified at the constraint.
4.4. A program was written to query the corporate MRP database on a
daily basis, which provided daily buckets of demand, instead of weekly.
4.5 The transit days on the MPS were changed to reflect the new
buffers.
4.6 Machine loads for shop resources were updated daily.
4.7 The constraint was scheduled based on daily reports. The MRP and
additional queries scheduled other resources.
4.8. Used existing WIP and finished goods inventory as a time buffer
while implementing the plan. Scrapped obsolete inventory, sold some at
discounts by working with marketing reps. Industrial engineers used
some inventory through substitutions and modifications of items.
4.9. Standard cost accounting remains in place for external reporting.
However, internal reports focus on rate of product flow through

constraint areas, and adherence to daily machine load schedules.
Database queries are used to manage constraint buffers.
Results: First full month after implementation set shipping records,
as did the next month. Overtime was held low, and orders were shipped
earlier than scheduled. Non-CCR resources are idle for a day or more
each week. Lead time for some products was less than half industry
standard. Due date shipment performance was in high 90%. Shipments
reached highest level in history, with personnel near the low post
layoff levels. Shipping costs as a percentage of sales fell.
Operating expenses were down 17%, earnings were up 154%, ROE was up
132%, inventory turns were up 39%, while sales volume fell 2%. (Koziol)
Company 3: Produced several million units per year with 20 variations.
Reorder point (ROP) is used for scheduling. An Out-of-stock condition
was not permitted for any product. Demand was high, steady, and
growing.
Page 11
Problems: The company had lead times of more than a month. Annual
material moves cost about $100K. Volume was increasing rapidly. There
was wide variability in machine productivity, downtime, and set up
time.
JIT Ideas Used: 1) No warehousing of WIP; 2) WIP would be controlled
through visual inspection, operator communication, and empty totes. 3)
re-engineer the process to decrease the number and time of setups.
Departures from JIT: 1) Buffer stocks allowed on line and in
warehouse.
TOC Ideas Used: 1) Capacity analysis, 2) Buffers in front of two CCRs;
3) Schedule product sequence to minimize changeover costs on the major
CCR; 4) run CCR machine at full capacity; 5) additional setups on
non-CCR machines
MRP Ideas used: 1) lead-time offsetting for certain processes
Implementation: 1) involve all groups in the planning and evaluation
process and address their concerns, building agreement and
understanding over time. 2) Implement the new system only on the last
half of the production process, where WIP parts were large, and had
largest payback.
Results: A big success, with annual savings of several hundred thousand
dollars, mostly from fewer material moves and warehouse savings.
Responsiveness was greatly improved. Production schedule lead times
were reduced to a few days. Finally, no capital expenditures were
necessary. (Giauque & Sawaya)
Other Results
Searching literature and the customers of the Avraham Goldratt
Institute (the leader vendor of TOC services) did not turn up any
examples of negative TOC results. However, Newman and Sridharan
surveyed 185 companies and found 5% using OPT based systems. Of these

5%, about half had less than spectacular results, with the worst
performances in the job shop environment. The best OPT performers were
in the process industries. The authors note that only about half of
the OPT users reported that the system had been in operation for more
than one year and that it may be too early to make any definite
conclusions. (Newman & Sridharan)
Page 12
APPENDIX
The Nine OPT rules
(1) balance flow, not capacity This is rapidly done through the DBR
heuristic.
(2) Constraints determine non-CCR utilization.
Maximizing utilization of non-CCR resources is not a goal, since
throughput is not affected, inventory goes up, and operating cost goes
up.
(3) Activation is not always equal to utilization
In other words, to activate a resource, when the output cannot go
through the CCR is wasteful in the form of excessive inventory. No
throughput is added for the additional inventory.
(4) An hour lost at a CCR is an hour lost for the entire system.
The CCR determines the factory capacity. To lose time here loses
throughput.
(5) An hour saved at a non-CCR is a mirage.
(6) CCRs govern throughput and inventory.
A scarce resource governs the other resources and thus the output of
the system. This is in contrast with MRP, where orders govern
inventory and throughput. The orders are the drum. If you do not let
the drum govern, there is no synchronization.
(7) Transfer batch should not always equal a process batch.
(8) Process batches should be variable, not fixed.

(9) Set the schedule by examining all the constraints simultaneously.
This is equivalent to solving a linear program. This approach also
avoids sub-optimization.
The theory of constraints has five steps:
Set up the system's goal and use the right measures.
(1) Identify the system constraint(s) (2) Decide how to exploit the
system constraints(s). (3) Subordinate everything else to the above
decision (4) Elevate the system constraint(s) (5) If, in the
previous steps, the constraint has been violated, go back to step 1,
but do not let inertia become the system constraint.
Page 13
BIBLIOGRAPHY
1. Bakke, Nils Arne, & Hellberg, Rolland, "Relevance lost? Acritical
discussion of different cost accounting principles in connection with
decision making for both short and long term production scheduling",
International Journal of Production Economics, 24 (1991) 1-18
2. Black, George, "U.K. Manufacturers Adopting JIT Technology Just In
Time", Software Magazine, October 1993, pp 111
3. Blodgett, Bill, & Laird, Bob, "WMJ Interview", Whirlpool Management
Journal, May 1993, pp 29-36
4. Couglan, Pamela, Darlington, John, "As fast as the slowest
operation: the theory of constraints" Management Accounting, June
1993, pp 14-17
5. Driscoll, J. & Hurley, S. "The development of a generic bottleneck
scheduler", Production Research: Approaching the 21st Century, Pridham,
Mark & O'brien, Christopher, Taylor & Francis, Inc. 1991
6. Elmes, David G. "The Thought Process Compared to Standard Theories
of Problem Solving: Old Wine in a New Bottle or a New Varietal?",
Department of Psychology, Washington and Lee University, August 1992
7. Gardiner, Stanley C, Blackstone Jr, John H., & Gardiner,
Lorraine R., "Drum-Buffer-Rope and buffer Management: Impact on
Production Management Study and Practices", International Journal of
Operations & Production Management, Vol. 13 No. 6, 1993, pp. 68-78

8. Goldratt, Eliahu, & Cox, Jeff, The Goal: A Process of Ongoing
Improvement, revised edition, North River Press, Croton-on- Hudson, NY
1987
9. ________________, & Fox, Robert E., The Race, North River Press,
Croton-on-Hudson, NY 1986
10. ________________, The Haystack Syndrome, North River Press,
Croton-on-Hudson, NY, 1990
11. ________________, The Theory of Constraints, North River Press,
Croton-on-Hudson, NY, 1990
12. Frizelle, G.D.M., "OPT in perspective", Advanced Manufacturing
Engineering, Vol 1, January 1989
13. Giauque, William C., Sawaya, William J. "Strategies for Production
Control", Production and Inventory management Journal--3Q92 pp 36-39
14. Inglesby, Tom, "MRP? Finite Scheduling? Yes, No & Maybe",
Manufacturing Systems, March 1991, pp 57-60
15. Jacobs, F. Robert, "OPT Uncovered: Many Production Planning and
Scheduling Concepts Can be Applied With Or Without The Software",
Industrial Engineering
16. Jayson, Susan, "Goldratt & Fox: Revolutionizing the Factory Floor",
Management Accounting, May 1987, pp 18-22
Page 14
17. Koziol, David S., "How the Constraint Theory Improved A Job- Shop
Operation"
18. Lambrecht, Marc R., & Decaluwe, Lieve, "JIT and Constraint Theory:
The Issue of Bottleneck Management" "Production and Inventory Managment
Journal, 3rd Quarter, 1988 pp 61-66
19. Lee, Chong Y., "A Recent Development of the Integrated
Manufacturing System: A Hybrid of MRP and JIT", International Journal
of Operations & Production Management, Vol 13, No. 4. 1993, pp 3-17
20. Neely, A.D., & Byrne, M.D. "A simulation study of bottleneck
scheduling", International Journal of Production Economics, 26, 1992,
pp. 187-192
21. Newman, William, Sridharan, V. "Manufacturing Planning and control:
Is there one definitive answer?", Production and Inventory Management
Journal, 1Q92, pp 50-53
22. Pinto, Peter A., & Rao, B. M., "Joint lot-sizing and scheduling for
multi-stage multi-product flow shops", Department of Management and
Department of Applied Statistics and Operations Research, Bowling Green
State University, pp. 1137- 1149

23. Plenert, Gerhard, & Best, Thomas D., "MRP, JIT, AND OPT: WHAT's
"BEST"?", Production and Inventory Management, 2Q 1986, pp 22-29
24. Plenert, Gerhard, "MRP, JIT, OPT: A Study", Modern Production
Management Systems, Elsevier Science Publishers B.V. (North-Holland),
1987, pp 729-734
25. Ptak, Carol A., "MRP, MRP II, OPT, JIT, and CIM--Succession,
Evolution, or Necessary Combination", Production and Inventory
Management Journal--2Q 1991
26. Ramsay, Martin L., Brown, Steve, and Tabibzadeh, Kambiz, "Push,
Pull and Squeeze Shop Floor Control With Computer Simulation",
Industrial Engineering
27. Ronen, Boaz, and Starr, Martin K., "Synchronized Manufacturing as
in OPT: From Practice to Theory" Computers ind. Engineering, Vol. 18,
No. 4, pp 585-600, 1990
28. Schonberger, Richard J., World Class Manufacturing, The Free Press,
NY, 1986
29. Schragenheim, Eli, & Ronen, Boaz, "Buffer Management: A
Diagnostic Tool for Production Control", Production and Inventory
Management Journal, 2Q91, pp 74-79
30. Sheridan, John H., "Throughput with a Capital 'T'", Industry Week,
March 4, 1991
Page 15

31. Spencer, Michael S., "Developing Finite Schedules for Cellular
Manufacturing", Production and Inventory management
Journal, 1Q, 1988, pp 74-79
32. ___________________, "The JIT, MRP, OPT Choice: How to CIM with
the Sharks", 33rd International Conference Proceedings,
APICS, 1990.
33. ___________________, "Using "The Goal" in an MRP System",
Production and Inventory management Journal, 4Q, 1991, pp 22-28
34. Vollmann, T.E., Berry, W.L., Whybark, D.C., Manufacturing Planning
and Control Systems, Richard D. Irwin Inc., 1988
35. Weeda, P.J., "Maximizing throughput by evaluating critical
utilization paths", International Journal of Production Economics, 24,
1991, pp 191-198
36. ___________, "On the relation between batch mode, bottleneck
machines and maximum throughput" pp 283-289
Page 16

TOC and MRPII

Recommended

Recommended

More Related Content

Similar to TOC and MRPII

Similar to TOC and MRPII (20)

TOC and MRPII