Note: Descriptions are shown in the official language in which they were submitted.
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
METHOD AND SYSTEM FOR CONTROLLING PROCESSES
TECHNICAL FIELD.
This invention concerns the control of processes, especially manufacturing
processes.
s BACKGROUND ART
A typical industrial manufacturing plant involves a number of processes which
produce products. Each process consists of a number of sub-processes. A
primary goal of plant management is to operate the plant as efficiently (as
close
to optimum performance) as possible to give the lowest product cost. However,
it
~o is not unusual for a plant to operate less than optimally. This may happen
for a
variety of reasons. Difficulties may be caused by one or more problems in the
process or sub-processes. For example, a power failure may stop all processes.
A jam in a cap chute in a bottling plant may cause un-capped bottles which may
require an extra operator to cap the bottles manually. In today's complex
~s manufacturing plants, there may be thousands of separate problems which
cause performance to deviate from optimum levels. Since resources for fixing
such problems are not unlimited, management is often faced with difficult
choices
as to which problems to solve and the order in which they should be solved.
Using the technology of the prior art, it was possible to track the
performance of
zo an industrial process by means of the following:
calculating the efficiency of the process (i.e. the ratio of the actual output
to the true potential output);
2. detem~ining reasons for downtime (i.e. the identity and duration of the
problems that stop the bottleneck);
zs 3. determining tabor variance (i.e. the difference between the actual
direct
labor cost and budgeted or planned direct labor cost);
4. determining raw material variance (i.e. the difference between the actual
CA 02321283 2000-08-17
PCT/AU99/00098
o:vdocslpatents\pct18412.doc Received 5 June 2000
2
raw material consumption and budgeted or planned raw material
consumption); and
5. determining the scrap level (i.e. the amount of product rejected during or
upon completion of the manufacturing process).
Each of these approaches has deficiencies. In particular, none of the
approaches of the prior art attempts to calculate the true financial cost of
an
individual problem nor the total cost of all problems affecting an industrial
process. It is therefore impossible for plant management to allocate resources
in
an efficient manner if prior art approaches are employed.
,o The present invention addresses these and other numerous deficiencies of
the
prior art.
DISCLOSURE OF THE INVENTION
The present invention comprises a method and a system for controlling a
process
for producing product, the process being designed to run at an optimum
~s performance, the method comprising the steps of:
identifying a problem occurring at some point in the process, which
problem causes the process to run at less than the optimum performance;
identifying a bottleneck in the process, such that the bottleneck
dictates the maximum speed at which the process runs;
zo determining the impact of the problem on the performance of the
process at the bottleneck;
calculating a financial value of the problem taking into account the
impact of the problem on the performance of the process at the
bottleneck;
is prioritising the problem based on the financial value of the problem;
and
AMENDED 8HE~
IpEAIAU
CA 02321283 2000-08-17
PCT/AU99/00098
Received 5 June 2000
o:ldocs~patents~pct~8412.doc
2a
adjusting the parameters of the process on the basis of the priority
of the problem determined by the prioritising step.
Preferably, the method further comprises the steps of determining an available
s optimum processing time and determining how much of the available optimum
processing time is lost due to the problem.
The amount of processing time lost due to a problem may be ascertained by
determining an available optimum processing time (or throughput) derived from
how much processing time (or throughput) would be available to the process if
1o the process were to run at the optimum performance and how much of the
product could be sold if made in the processing time which would be available
to
AMENDEd SH~C'
~pEpJAU
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
3
the process if the process were to run at the optimum performance. The
available optimum processing time and therefore the amount of processing time
lost due to the problem thus reflect market conditions for the product made by
the
process.
In one embodiment, the invention is capable of determining the true financial
cost
of problems affecting the process, thus enabling the different problems to be
prioritised and appropriate resources deployed for each problem. A typical
process for producing a product is designed to run at an optimum performance.
However, the process may have one or more problems which cause the process
1o to run at less than the optimum performance, thereby causing the process to
lose
processing time or throughput.
In another embodiment, the method of the invention includes identifying a
problem in the process and identifying a bottleneck in the process, such that
the
bottleneck dictates the maximum speed at which the process runs. A financial
~s value of the problem is determined, taking into account the impact of the
problem
on the bottleneck. The probtem can then be prioritised, a decision can be made
whether to correct the problem or resources can be allocated to the problem
based on the financial value of the problem.
Another aspect of the method of the invention is valuing the problem based on
zo marginal proFtability of the product, thus reflecting market conditions and
allowing
market conditions to be used in prioritising. the problem. As another option,
the
financial value of the problem may be determined by valuing the processing
time
the process loses due to the problem based on labour cost.
Another aspect of the method of the invention is determining how much
25 processing time the process loses due to the problem. The step of
calculating the
financial value of the problem may comprise the step of valuing the problem
based on how much processing time the process loses due to the problem.
A process may be production constrained or sales constrained. The process is
production constrained when as much of the product as the process can produce
3o can be sold in the market. The process is sales constrained when not all
the
CA 02321283 2000-08-17
WO 99/42939 PCTIAU99/00098
4
product that can be produced can be sold. Another aspect of the invention is
the
determination of whether the process is production constrained, and the
determination of whether the process is sales constrained in order to
determine
the financial value of the problem. Yet another aspect of the invention is the
s determination of the extent to which the process is production constrained,
and,
to the extent that the process is production constrained, valuing the
processing
time the process loses due to the problem based on marginal profitability of
the
product. Still another aspect of the invention is valuing lost process time
based
on labour cost to the extent that any lost process time is not valued based on
to marginal profitability.
Yet another aspect of the method of the invention is determining how much
processing time the process loses due to the problem causing the bottleneck to
stop running, how much processing time the process loses due to the problem
causing the bottleneck to run slower than expected and how much processing
~s time the process loses due to the problem causing product to be scrapped at
or
after the bottleneck.
The present invention also comprises a computer system for performing the
method of the invention. The computer system of the present invention may
include means for inputting modeling data relating to the process, thereby
2o building a computer model of the process and means for inputting
performance
data obtained by monitoring performance of the process. In one embodiment,
means are provided for identifying a bottleneck in the process such that the
bottleneck limits the maximum speed at which the process runs. The system can
also include means for determining a financial value of problems in the
process
is which include means for determining an amount of processing time the
process
loses due to a problem or one of a plurality of problems by taking into
account
the impact of the problems on the bottleneck.
Another aspect of the invention is means for determining how much of the
product could be sold if made in the time which would be available if the
process
3o were to run at the optimum performance. The amount of product which could
be
sold is used to determine the available optimum processing time, thus
reflecting
CA 02321283 2000-08-17
PCT/AU99/00098
o:vdocsvpatentslpct18412.doc Received 5 June 2000
market conditions in the value of the problem.
Another aspect of the invention is means for valuing the available optimum
processing time lost due to the problem based on marginal profitability of the
product and means for valuing the available optimum processing time lost due
to
the problem based on labor cost.
Further aspects of the invention involve means for determining the extent to
which the process is production constrained and sales constrained; and means
for valuing the processing time the process loses due to the problem based on
marginal profitability of the product to the extent that the process is
production
,o constrained and based on labor cost to the extent that the process is sales
constrained.
Another aspect of the computer system is means for valuing problems based on
marginal profitability of the product, thus reflecting market conditions and
allowing
market conditions to be used in prioritising the problems.
~s Another aspect of the computer system is means for determining a slow
running
time for the problem, indicative of how much processing time the process loses
due to the problem causing the bottleneck to run at a speed slower than an
expected speed; means for determining a downtime for the problem, indicative
of
how much processing time the process loses due to the problem causing the
2o bottleneck to stop running; and means for determining a bottleneck waste
time
for the problem, indicative of how much processing time the process loses due
to
the problem causing product to be scrapped at or after the bottleneck. Means
are
provided for valuing the problem based on the amount of the processing time
lost
due to the problem.
is Preferably, the method further comprises the step of determining how much
additional crew operates the process due to the problem, wherein the step of
determining the financial values comprises determining the financial value of
how
much additional crew operates the process due to the problem, and the step of
determining how much the problem reduces profitability is also based on the
so financial values of how much additional crew operates the process due to
the
AMENDED SHEE r
IPEAIAU
CA 02321283 2000-08-17
_ PCT/AU99/00098
o:vdocs~patents~pct~s412.doc Received 5 Jane 2000
5a
problem.
Preferably, the method yet further comprises the steps of determining how much
_
raw material the process uses due to the problem, wherein the step of
determining the financial values comprises determining comprises determining
the financial value of how much raw material the process uses due to the
problem, and the step of determing how much the problem reduces profitability
is
also based on the financial values of how much raw material the process uses
due to the problem.
These and other aspects of the invention will be apparent from the Brief
~o Description of the Drawings and Claims.
BRIEF DESCRIPTION OF THE DRAWINGS
Fig. 1 is a block diagram showing the present invention applied to a simple
potato processing and packaging plant;
AMENDED SHEE1
IPEAlAU
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
s
Fig. 2 is a flowchart showing the basic steps involved in the present
invention;
Fig. 3 is a snapshot of the main computer screen for the computer program of
the
present invention;
Fig. 4 is a snapshot of the computer screen used for entry of shift data for
the
s modeling of the process to be controlled using the present invention;
Fig. 5 is a snapshot of the navigation menu bar used for the entry of process,
variant, sub-process and raw material data for the modeling of the process to
be
controlled using the present invention;
Fig. 6 is a snapshot of the computer screen used for entry of process data for
the
~o modeling of the process to be controlled using the present invention;
Fig. 7 is a snapshot of the computer screen used for entry of variant data for
the
modeling of the process to be controlled using the present invention;
Fig. 8 is a snapshot of the computer screen used for entry of sub-process data
for the modeling of the process to be controlled using the present invention;
~s Fig. 9 is a snapshot of the computer screen used for entry of raw materials
data
for the modeling of the process to be controlled using the present invention;
Fig. 10 is a snapshot of the computer screen used for entry of reason sets for
the
modeling of the process to be controlled using the present invention;
Fig. 11 is a snapshot of the computer screen used for attaching reason sets to
zo variants;
Fig. 12 is a snapshot of the computer screen used for attaching reason sets to
sub-processes;
Fig. 13 is a snapshot of the computer screen used for the entry of quantity
sets;
Fig. 14 is a snapshot of the computer screen used for attaching reason sets to
is the quantity sets;
CA 02321283 2000-08-17
WO 99!42939 PCT/AU99/00098
7
Fig. 15 is a snapshot of the computer screen used for selecting process,
shift,
date and variant and also for entry of production data;
Fig. 16 is a snapshot of the computer screen used for entry of downtime data;
Fig. 17 is a snapshot of the computer screen used for entry of yield data;
s Fig. 18 is a snapshot of the computer screen used for entry of waste data;
Fig. 19 is a flow diagram showing the basic steps in the valuation of problems
and the control of a process in accordance with the value of problems;
Fig. 20 is a flowchart showing the calculation of the available production
constrained savings table over the base table period;
~o Fig. 21 is a flowchart showing the calculation of the value of lost labor
time
caused by a problem over the base table period;
Fig. 22 is a flowchart showing the lost production cost evaluation method;
Fig. 23 is a flowchart showing the calculation of the value of the downtime
component of a problem over the base table period;
~s Fig. 24 is a flowchart showing the calculation of the value of the slow
running
time component of a problem over the base table period;
Fig. 25 is a flowchart showing the calculation of the value of the bottleneck
waste
component of a problem over the base table period;
Fig. 26 is a flowchart showing the calculation of the value of the raw
material
Zo waste component of a problem over the base table period;
Fig. 27 is a flowchart showing the summing of the components of the value of a
problem over the base table period, the extrapolation of the value and the
prioritising of the problem based on its value.
DETAILED DESCRIPTION
25 The following is a detailed description of the invention. For ease of
CA 02321283 2000-08-17
WO 99/42939 PGT/AU99/00098
8
understanding, as aspects of the invention are described, reference is made to
an example of a simple industrial process for packaging potatoes. The
description and example are not intended to be limiting, the scope of the
invention being determined by the claims.
s A typical factory involves several shifts each day and several processes
producing "variants" from raw materials. Each process may have sub-processes.
A "variant" is a product or products whose characteristics are very similar.
In the potato processing and packaging example, broadly illustrated in Fig. 1,
factory 1 receives raw materials 2 and processes them by means of processes
(or lines) 3, 4 and 5 which produce variants 6, 7, 8 and 9 - cases of 1, 2 and
3 Ib
cans of potatoes and variants 10 and 11 - packets of salted and plain potato
chips. Each variant has a number of different characteristics. These include
the
optimum crew size for operating the process, the maximum bottleneck speed of
the process, the selling price of the variant, the unit marginal manufacturing
cost
~s of the variant and the names of various problem sets. A number of different
problems may arise in a process. "Problem sets" are problems of different
types
which cause the characteristics of the process to deviate from their optimum
or
expected values. For example, the process may run too slowly ("slow running"
problems), excess crew may be assigned to a process ("excess crew" problems),
the process may stop working ("down time" problems), units may be scrapped at
or after the bottleneck which wastes process time ("bottleneck waste"
problems)
and raw materials may be wasted ("raw material waste" problems).
The invention comprises a personal computer 12 programmed to model industrial
process or factory 1 and perform certain calculations using custom application
is 13. Once the calculations have been performed, a problem priority list is
printed
out by printer 14 and is used to prioritise the adjustment of the parameters
of
factory 1.
Custom application 13 comprises specially written program code which interacts
with a commercially available database 15, preferably Access~, available from
3o Microsoft. Custom application 13 and database 15 preferably run in a
Windows~
95 environment, but the precise operating system and database are not cnrcial
to
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99I00098
9
the invention. The custom application and its interactions with the database
will
now be described in detail, it being understood that it is within the
capabilities of
a person of ordinary skill in the art to write the computer code needed to
implement the invention.
s The most basic steps of the method of the invention are shown in Fig 2.
First,
the process or factory is modeled at S1. Next, process data is gathered and
input into the database at S2. Financial values are calculated for the
problems
and a problem priority table is generated at S3 based on those values. The
parameters of the process are adjusted based on the problem priority table at
S4.
Each of these steps will now be described in detail.
1. MODELING THE FACTORY OR PROCESS:
In order to model the factory or process, data is entered into the computer
via the
main menu screen 20 of custom application 13 shown in Fig. 3. The user clicks
the mouse or pointing device of computer 12 on the Tasks icon 24 in the menu
~s bar 22, pulls down a menu 2fi and clicks on the Model Factory icon 28. That
takes himlher into the Factory Model screen 30 shown in Fig. 4. As shown in
Fig.
4, the Factory Model screen is made up of a graphical representation of ten
index
cards, in a configuration commonly used in Windows 95 applications. Each of
the index cards contains fields for entering data or drop down lists
containing
Zo data already entered. These are used to build the mode! of the process or
factory.
The index cards are as follows: Shift Card 32, Process Card 34, Variant Card
36,
Raw Materials Card 38, Sub Process Card 40, Reason Sets Card 42, Quantity
Sets Card 44, Targets Card 46, Custom Measures Card 48 and Custom
Variables Card 50. Cards 32 to 44 will be described in detail. Cards 46, 48
and
50 have no application to the present invention.
Each screen displaying Process Card 34, Variant Card 36, Raw Materials Card
38, Sub Process Card 40 has a number of records. The user can navigate
through the records by means of the navigation menu bar 51 shown in Fig. 5.
3o The navigation menu bar has the following buttons: MoveNext 53, which skips
to
CA 02321283 2000-08-17
WO 99/42939 PGT/AU99/00098
1U
the next record, MovePrevious 52, which skips back to the previous record,
MoveFirst 55, which skips to the first record and MoveLast 56 which skips to
the
last record. The record number is displayed in field 54. New records are added
by moving to the last record and pressing MoveNext. This gives a blank record.
s When the record is filled or edited, the user presses MoveNext to save the
information.
a. Shift Card:
Shift card 32 contains three fields 321, 322 and 323 which define the shifts
operating in the factory. In the potato processing and packaging example
~o illustrated in Fig. 4, there are three shifts: day, afternoon and night.
The user
simply enters the names of the shifts in the shift fields and then clicks on
the next
index card.
b. Process Card:
Process Card 34 shown in Fig. 6 permits entry of data for the various
processes
~s in the plant. It contains three fields: Process Name 341 which identifies
each
process, All Up Labor Rate ("AULR") 342 and Target Set 343. The All Up Labor
Rate entered at field 342 is the total hourly cost of employing operators for
a
particular process. The Target Set drop down list 343 is not used for the
present
invention. The Process Card also allows the user to set a flag 344 which
2o indicates whether a particular process is a bottleneck process, meaning
that it is
the limiting process in the factory. Flag 344 is not applicable to the present
invention.
In the potato processing and packaging example, there are three processes,
line
K1 for putting potatoes into 1 Ib cans, line K2 for putting potatoes into 1, 2
and 3
25 Ib cans and the Chipping Line which produces salted and plain potato chips.
The
data entered for the various records stored in database 15 via the Process
Card
for the potato processing and packaging example are shown in the following
table:
CA 02321283 2000-08-17
WO 99/42939 PGT/AU99/00098
11
_ ,-
Line: K1 K2 Chipping Line
Process Name K1 K2 Chipping Line
All Up Labor Rate20 20 22
Target Set NIA NIA NIA
Bottleneck Yes Yes Yes
Process?
TABLE 1: Process Records
c. Variants Card:
As stated above, a variant is a product or products whose characteristics are
very
s similar. Each variant must be defined in order to model the factory. In our
example, the variants are 1, 2 and 3 Ib cans of potatoes and packets of potato
chips. The characteristics of the variants of the process are entered at
Variant
Card 36 shown in Fig. 7. Each characteristic is entered into a specific field
in
Variant Card 36. The various fields and drop down lists are:
~ process name, for example, K1 for 1 Ib cans of potatoes - drop down list
361.
~ variant name, for example, 1 Ib cans - drop down list 362.
production units - for example cases of 1 Ib cans - field 363.
~ maximum bottleneck speed of variant "v" ("MBS~') - field 364 - this is the
maximum speed at which a process can run (i.e. the maximum speed of the
slowest sub-process) and is measured in production units per hour. Each
process has a bottleneck sub-process which dictates the maximum speed at
which the process can run. This must be identified in each case.
~ optimum crew size to produce variant v ("OCSY") - the optimum number of
human operators needed for running a particular process to produce variant v
zo - field 365.
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
12
~ unit marginal manufacturing cost for variant v ("UMC~") - the cost of
producing
an additional production unit of the variant, including raw materials,
packaging
materials, additional energy costs, additional waste removal costs and
additional distribution costs - field 366.
s ~ unit sates price of variant v ("USP~") - the sales price per production
unit - field
367.
~ production constrained additional volume of variant v ("PC%~") - the
percentage of additional volume of the variant which could be sold, over and
above the current level of production - field 368.
~ capital expenditure - field 369 - not used in the present invention.
~ ABS conversion name - drop down list 370 - not used in the present
invention.
~ Output conversion name - drop down lists 371 - not used in the present
invention.
~ Reason Sets - the possible reasons for problems with the process - drop
~s down lists 372, 373 and 374. The details of these will be described below.
~ Missed Plan Reason Set and Target Set - drop down lists 375 and 376 are
not used in the present invention.
~ In the potato processing and packaging example, the data for each variant
are shown in the following table:
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99100098
13
Process K1 K2 K2 K2 Chipping Chipping
Name: Line Line
Variant 1 Ib 1 Ib 2 Ib can 3 Ib can plain salted
Name: can can
Production cases cases cases cases pounds pounds
Units:
MBS: 1000 1000 600 400 3000 2000
OCS: 25 28 28 28 15 17
UMC: 2 2 5 7 .55 .6
USP: 5 5 9 12 2 2
PC%: 10 10 0 2 20 20
TABLE 2: Variant Records
d. Sub-Processes Card:
s Each process can be broken down into sub-processes. A sub-process is a
logical step or piece of equipment used in the process, typically a machine,
for
example, the canning machine for putting potatoes into cans or the salting
barrel
for salting the potato chips. While there may be a number of different sub-
processes in a process, only those which are relevant need to be identified
and
included in the factory model. Relevant sub-processes include:
~ The Speed Bottleneck Sub-Process: This is the sub-process with the lowest
maximum speed and hence the speed-limiting sub-process. This sub-process
must be modeled.
~ The Output Bottleneck Sub-Process: This is the sub-process which limits the
~s output of the process for the majority of the time. It may be the same as
the
Speed Bottleneck Sub-Process. The Output Bottleneck Sub-Process is the
sub-process with the lowest product of maximum bottleneck speed and
percentage up time. Modeling this process is optional. .
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
14
~ Near Bottleneck Sub-Processes: These are processes which are close to
being the output bottleneck or often are. Modeling these processes is
optional.
Reverting to the potato packing process, the speed bottleneck sub-process for
s the 1 Ib can process K1 is the can filler. It is also the output bottleneck.
The
same is found to be case for the 2 Ib process, K2. in the chipping line, the
speed
bottleneck is the fryer for plain chips and the salting barrel for salted
chips. For
the sake of simplicity, in this example, it will be assumed that there are no
other
sub-processes.
~o Modeling each process includes entering relevant sub-process information at
Sub-Process Card 40 shown in Fig. 8. The user selects the process (e.g. the
chipping line) by means of drop down list 401, and the variant made by that
process (e.g. salted chips) by means of drop down field 402. The sub-process
(e.g. the satting barrel - the speed bottleneck sub-process in this case) is
entered
at drop down list 403 or can be selected from a list of sub-process names. The
downtime reason set name is displayed in drop down list 404, which contains a
list of the reason sets (reasons for down time at the particular sub-process
entered as described in the text relating to Fig. 10). The Percentage of
Bottleneck is entered at field 405. This is the percentage of throughput
through
zo the particular bottleneck sub-process. For example, if two parallel can
fillers
formed a bottleneck, each filler would account for a proportion of the
bottleneck.
In the eXample illustrated in Fig. 8, the salting barrel sub-process
represents
100% of the bottleneck throughput.
The data for the speed bottleneck sub-process for each of the processes K1, K2
25 and the chipping processes are shown the following table:
CA 02321283 2000-08-17
WO 99/42939 PGT/AU99/00098
Process K1 K2 K2 K2 Chipping Chipping
Name: Line Line
Variant 1 Ib 11b can 2 Ib can 3 Ib can plain salted
can
Name:
Sub-Process K1 FillerK2 FillerK2 FillerK2 Fillerfryer salting
Name: barrel
Of 100 100 100 100 100 ~ ~ 100
Bottleneck:
TABLE 3: Sub-Process Records
e. Raw Materials Card:
s Data relevant to the value of raw materials is entered into the model at Raw
Materials Card 38 shown in Fig. 9. For each variant, there is at least one raw
material. Raw materials are defined in terms of:
~ the process name - drop down list 381 - containing the names of all the
processes (see Fig. 6).
~o ~ the variant name - drop down list 382 - containing the names of all the
defined
variants (see Fig. 7).
~ the raw material name - drop down list 383, allowing direct entry or
selection
from previously defined raw material names.
~ the units of raw material - drop down list 384 - containing the names of all
~s reason quantity sets (see Fig. 13).
~ the per unit cost of raw material (RW$~) - field 385.
~ the minimum raw material content - field 386 - this is the minimum quantity
of
raw material required to produce one production unit of the variant, assuming
no waste loss.
CA 02321283 2000-08-17
WO 99/42939 PGT/AU99/00098
16
The target set - drop down list 388 - and conversion name - drop down list 389
-
are not pertinent to the present invention.
The data for the raw materials used in processes K1, K2 and the chipping lines
are shown in the following table:
Process Variant Raw Raw Raw Cost Minimum Raw
Name Name Material Units Material
Name Content
K1 1 Ib cans potatoes Ibs 0.1 10
K1 1 Ib cans cans cans 0.05 10
K2 1 Ib cans potatoes ibs 0.1 10
K2 1 Ib cans cans cans 0.05 10
K2 2 Ib cans potatoes Ibs 0.1 20
K2 2 Ib cans cans cans 0.1 10
K2 3 Ib cans potatoes Ibs 0.1 30
K2 a lb cans cans 0.15 10
Chipping plain potatoes Ibs 0.1 3
Line
Chipping plain bags bags 0.02 1
Line
Chipping salted potatoes Ibs 0.1 3
Line
Chipping salted bags bags 0.02 1
Line
TABLE 4: Raw Material Records
f. Reason Sets Card:
A "Reason Set" is a group of reasons for different problems in the factory or
process (excess crew problems, downtime problems, slow running problems,
~o bottleneck waste problems and raw material waste problems). Reason sets are
defined by inputting information at the Reason Sets Index Card 42 shown in
Fig.
10. Each reason set has a name shown in drop down list 421. The reason set is
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
17
entered into drop down list 421. Each reason set consists of one or more
reasons for a particular problem. In the illustrated example, the K1IK2
Downtime
set comprises changeovers, cans jammed at the depalletizer, jams at the infeed
to the filler etc. as shown in table entered at table 422. These are entered,
one
s reason definition per line. The columns 423 are the characteristics of each
reason. When the operator clicks on a particular reason set in the pull down
menu at 421, the table of reasons in that reason set and the corresponding
characteristics columns appear. The relevant characteristic data can then be
entered. The properties of the illustrated reason set are:
~ Reason Code - column 424 - this is a number indicating the order of a
particular reason within the reason set.
~ Reason Name - column 425 - this is a descriptive title for a reason in the
reason set, for example, the reason set shown in Fig. 10 includes a reason
name "changeovers." This indicates that a problem is caused by
~s changeovers between variants.
~ Bottleneck Waste Flag - column 426 - this is a flag which signifies that a
particular reason in the reason set acts as a bottleneck waste reason, in
addition to any other type of reason that it might be.
~ Waste Reporting Unit ("WRU~;') - column 427 - these are the units of waste
Zo in which the particular reason wiH be reported, for example, kilograms of
raw
potatoes, cans, or cases.
~ Waste Reporting Units to Production Units ("WRUtoPU~,") - field 428 - this
is
the numerical conversion factor to convert the waste reporting units to
production units.
zs Note that the following fields and flags shown in index Card 42 are
inapplicable
to the present invention: Setup flag, subset name, DT category flag, Base DT
category, Waste Entry Conversion Code, DT target, SR target, XC target and BW
target.
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
18
Four reason sets are defined in the example which has been used to illustrate
the modeling of the factory. The data entered into the Reason Set Card for the
potato processing and packaging plant are tabulated in Tables 6, 7, 8 and 9
for
K1/K2 waste, K1/K2 slow running, K1/K2 downtime and K11K2 excess crew
s respectively.
Reason Code Reason Name Bottleneck Waste Waste
Waste Flag Reporting Reporting
Unit Unit to
Production
Units
1 changeovers true full cans 0.1 (i.e.
10
cans per case)
2 cans falling true full cans 0.1
from conveyor
3 jams at infeedfalse cans
to filler
4 cans jammed false cans
at depalletizer
leakage true full cans 0.1
TABLE 6: K11K2 Waste Reason Set
CA 02321283 2000-08-17
WO 99/42939 pCTIAU99100098
19
Reason Code Reason Name Bottleneck Waste Waste
Waste Plag Reporting Reporting
Unit Unit to
Production
Units
1 undefills false
2 faulty seal false
on
fail heads
3 leakage false
4 operator false
choice
wrong setting false
after
changeover
TABLE 7: K11K2 Siow Running Reason Set
Reason Code Reason Name Bottleneck Waste Waste
Waste Flag Reporting Reporting
Unit Unit to
Production
Units
1 changeovers false
2 cans jammed false
at depalletizer
3 jams at infeedfalse
to filler
4 leakers false
5 spill at fillerfalse
outfeed
6 faulty seal false
on
f Iler
s TABLE 8: K1IK2 Downtime Reason Set
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99100098
Reason Code Reason Name Bottleneck Waste Waste
Waste Flag Reporting Reporting
Unit Unit to
Production
Units
1 no label false
applied
2 leakage false
3 too many false
temps booked
4 forced productfalse
changeover
TABLE 9: K11K2 Excess Crew Reason Set
g. Attaching Reason Sets:
s Having entered the data defining the various reason sets, the reason sets
can be
attached to the variants and subprocesses. It will be recalled that when
Variants
Index Card 36 was described (see Fig. 7), the Reason Set fields were not
discussed. The Reason Sets fields in Fig. 7 (i.e. Excess Crew 372, Slow
Running 373, Bottleneck Waste 374 and Raw Material Waste 376) are pull down
~o menus which allow the user to select a known Reason Set entered at the
Reasons Sets Card. The attachment of Reason Sets to variants is illustrated in
Fig. 11. There are similar reason sets for the sub-processes which are
similarly
attached by means of the sub-process screen as shown in Fig. 12.
h. Quantity Sets:
Waste is typically reported in terms of units of finished or partially
finished
product (e.g. cans of potatoes or cases of cans). It is therefore necessary to
provide a conversion factor "WRUtoRU~" befinreen waste reporting units
("WRU~,") and raw material wasted("WasteQty~",a "). The definition of Reason
Quantity Sets provides this conversion factor.
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99100098
21
A reason quantity set must be defined between each reason set which is to be
used for raw material waste problems and each raw material. In the case of the
potato processing and packaging plant, there are two raw materials for each
production variant. A reason quantity set will be required for each raw
material
s for each variant, unless the data are identical.
The user clicks on the Quantity Sets index card 44. Refer now to Fig. 13. Drop
down list 441 contains the names of all the reason sets already defined (see
Fig.
10). The name of the quantity set is entered at field 442. Table 443 is made
up
of each reason in the reason set which is specified in field 441. The waste
reporting unit (column 444) comes from the definition set up in Fig. 10
(column
427). The only data entered is WRUtoRU~,, entered in column 445.
Four Waste Reason Quantity Sets are required for the K1 and K2 lines of the
potato processing and packaging plant described. One for cans (i.e. one for
21b
and one for 31b cans) and two are for potatoes. These are shown in Tables 10,
~s 11 and 12.
Reason Reason Name Waste Reporting Waste Reporting
Code Unit Unit
to Raw Material
Units
1 changeovers full cans 1
2 cans fallen from full cans 1
conveyor
3 jams at infeed cans 1
to filler
4 cans jammed at cans 1
depalletizer
leakage full cans 1
TABLE 10: K1IK2 Cans Reason Quantity Set
CA 02321283 2000-08-17
WO 99142939 PGT/AU99I00098
22
Reason Reason Name Waste Reporting Waste Reporting Unit
Code Unit to Raw Material Units
1 changeovers full cans 2
2 cans fallen from full cans 2
conveyor
3 jams at infeed cans 2
to filler
4 cans jammed at cans 2
depalletizer
leakage full cans 2
TABLE 11: K1IK2 2lb Potatoes Reason Quantity Set
Reason Reason Name Waste Reporting Waste Reporting Unit
Code Unit to Raw Material Units
1 changeovers full cans 3
2 cans fallen from full cans 3
conveyor
3 jams at infeed cans 3
to filler
4 cans jammed at cans 3
depailetizer
5 leakage full cans 3
s TABLE 12: K11K2 Cans Reason Quantity Set
i. Attaching Reason Quantity Sets to the Factory Model:
Refer to Fig. 14. The raw materials card 38 is used to attach the reason
quantity
sets entered at the quantity sets screen, just described, to the model. The
waste
reason quantity set is selected using drop down list 387 and attached by means
~o of the navigation menu bar.
CA 02321283 2000-08-17
WO 99/42939 PGT/AU99/00098
23
2. INPUT OF PROCESS DATA:
Having created a model of the process or factory, the next step is the entry
of
data relating to the actual performance of the process or factory. For each
production run, factory personnel will enter the data for each variant. There
are
s four categories of data: (1 ) production data; (2) downtime data; (3) yield
data and
(4) waste data.
On the main screen 20, the operator clicks the pointing device on Task 24 in
the
menu bar 22, pulling down the task menu 26. See Fig. 3. He/she then clicks on
the Production Data Entry icon 25. This brings up the Data Entry screen 60
shown in Fig. 15. The operator then selects the shift (e.g. afternoon) in drop
down list 601, enters the date {e.g. 1I1 ) in field 602, selects the process
{e.g. the
K1 fine) in drop down list 603 and the variant (e.g. 11b cans) in drop down
list
604. For each of the categories of data (1 ) production data; (2) downtime
data;
{3) yield data and (4) waste data, there is a corresponding index card 610,
620,
~s 630 and 640 respectively.
Clicking on the Production Index Card 610 {see Fig. 15) allows the operator to
enter production data. In each case, the operator will enter the actual number
of
hours the process was crewed for the relevant shift ("APH"~") (e.g. 8 hours)
in
field 611, the speed of the process - the actual bottleneck speed in units per
hour
Zo ("ABS"~ ")(e.g. 800) in field 612, the output in units ("O~~")(e.g. 5,000)
in field 613
and the actual crew size ("ACS,~d")(e.g. _25) in field 614. The operator then
uses
the drop down lists 615 and 616 to choose a reason for excess crew and a
reason for slow running. See Fig. 7, fields 372 and 3T3.
Referring to Fig. 16, the operator clicks on Downtime Index Card 620 to enter
25 data relating to downtime and number of stops in a particular production
run by
variant. For the same shift, date, process, and variant just selected, a
number of
downtime entries can be entered. Downtime index Card 620 includes table 621
for that purpose. Table 621 has a sub-process drop down list 622 which allows
selection of a sub-process from the set of sub-processes defined in the sub-
so processes screen 40 {See Fig. 8), for the selected process and variant. The
appropriate reason set for the sub-process is then made available for the set
of
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
24
downtime reasons. The user then selects a reason code from drop down list 623
(entered at the Reasons Sets screen 42) (See Fig. 10) and attached to the
particular sub-process in the sub-process screen 40 (See Fig. 12). Having
selected a sub-process and a reason code, the operator can now enter the
s downtime ("DTI") and the number of stops in fields 624 and 625 respectively.
This is repeated for all downtime which has occurred on atl sub-processes on
the
selected shift for the selected variant.
Fig. 17 shows the Yield Index Card 630. Clicking on this card allows entry of
the
quantity of raw materials which were used for the selected process. For the
~o same shift, date, process, and variant just selected, a defined raw
material ("raw
name") for a particular variant can be selected by clicking on the drop down
list
632 in table 630. This list contains all the raw materials for the selected
variant.
This was defined in the Raw Materials screen (see Fig. 9). The quantity of
each
raw material ("RI"sd ") used in the production run of the selected variant on
the
~s selected shift is entered in column 632.
Fig. 18 shows the Waste Data Index Card 640. Clicking on this card allows the
entry of data for raw material waste and bottleneck waste. For the same shift,
date, process, and variant just selected, a waste reason set is selected (see
Fig.
7, field 374) by clicking on the code drop down list 64, which lists all
defined
Zo codes previously entered (see Fig. 10). Once the reason code has been
selected, the waste reason description is automatically loaded. See column
642.
The quantity of raw material wasted ("WasteQty~") is entered in column 643
and the frequency of waste is entered in column 644. This is repeated for all
raw
material and waste problems that occurred during production of the variant on
is the particular day and shift.
3. GENERATING THE PROBLEM PRIORITY TABLE FOR PROBLEMS IN
A PROCESS:
The factory or process has now been modeled and the relevant performance
data have been entered. The operator now clicks on the print reports icon 27
in
ao main screen 20 (see Fig. 3). This starts the calculation of the "Problem
Priority
Table" and its printing.
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
The following is a detailed description of the calculation of the problem
priority
table. Refen-ing to Fig. 19, there are ten basic steps in the development of
the
problem priority table, namely:
1. Building the "Available Production Constrained Savings Table." Step S10
s 2. Valuing the excess crewing component of a problem p over the base table
period, BTP (XCLL$~,). Step S12. The base table period is typically five
weeks.
3. Valuing the process downtime component of a problem p over the BTP
(DT$~,). Step S14. '
~0 4. Valuing the process slow running time component of a problem p over the
BTP (SR$~"). Step S16.
5. Valuing the process time lost due to bottleneck waste (the product
rejected at or after the bottleneck sub-process) component of a problem p
over the BTP (BW$~,). Step S18.
~s 6. Valuing the wasted raw material component of a problem p over the BTP
(RW$~,). Step S20.
7. Summing the total value of problem p over the BTP. Step S22.
8. Extrapolating the cost of problem p over a one year period. Step S24
9. Repeating steps 1 to 8 for all problems occurring in the process over the
2o BTP. Steps S26 and S28.
10. Sorting by total value all problems to generate the problem priority
report,
printing out the report and adjusting the process in accordance with the
value of the problems. Step S30.
11. Each of these steps will now be described in detail.
CA 02321283 2000-08-17
WO 99/42939 PGT/AU99/00098
26
1. Building the "Available Production Constrained Savings Table."
If the process is not operating at 100% efficiency, bringing it to '! 00%
efficiency
will allow additional units of the variant to be made. If those units can be
sold
(i.e. the process is "production constrained") then additional profits will be
s derived. If those units cannot be sold (i.e. the process is sales
constrained), then
improving the efficiency of the process will result in production of the same
number of units in less time, thus reducing the labor costs.
The Available Production Constrained Savings Table analyses the process in
terms of time units of 100% efficient operation, i.e. minutes during which the
,o process is performing as expected over a base table period ("BTP"). The
base
table period is preferably 5 weeks. Once that is done, each problem can be
valued in terms of how many of such units it wastes, in addition to how much
raw
material and labor time it wastes.
As a first step, the number of units of 100% efficient operation (or "optimum
process time") available, PCM~, for the process over the BTP and the monetary
value of those units, PCM$~, are calculated. The Available Production
Constrained Savings Table, stored in database 15, identifies each variant for
which additional volume could be sold over the BTP (i.e. each production
constrained variant), the quantity of optimum process time required to produce
Zo the additional volume, PCMinsr", and the value of each minute of that
process
time, PCMins$~,. These values will be used in later steps in osier to value
the
portion of the problem relating to the cost of lost production, i.e. the lost
profit that
could have been generated from the sale of additional units.
Refer to Fig. 20. PCU~" the quantity of additional volume of a variant v which
25 could be sold over the BTP is calculated as follows. The output of variant
v over
the BTP, 0~,, and the percentage of units which could potentially be sold if
they
were made, PC%~ - expressed as a percentage of current production, are
obtained from database 15. PCU~, is found by multiplying PC%" by Om. Step
S40.
3o Next, PCU$~" the value (total profit) of the additional units of variant v
which
CA 02321283 2000-08-17
WO 99/42939 PCTIAU99100098
27
could be sold, is calculated by multiplying PCU~, by the unit marginal
profitability
UMP~ of variant v. Step S42. The unit marginal profitability, UMP~, calculated
at
Step S44, is the difference between the unit sales price of a variant, USP",
and
the unit marginal manufacturing cost, UMC~, both of which are stored in
database
s 15.
PCMins~" the optimum process time required to produce the additional variants
is then calculated by dividing the quantity of additional variants which~could
be
sold, PCU~" by the maximum bottleneck speed for the particular variant, MBSy
(stored in database 15). Step S46.
~o Finally, PCMins$~" the value of each optimum process minute, if used to
produce
the additional sales volume of variant v, is calculated. This is done by
dividing
the value of the additional units of variant v, PCU$~" by the optimum process
time required to produce those additional variants, PCMins~,. Step S48.
These steps are repeated for all variants which were produced in the base
table
~s period and the Available Production Constrained Savings Table is built.
Step
S50. The variants are arranged in the table shown in Step S50 in descending
order by PCMins$~,, the value of each optimum process minute. The table has
tour columns, namely variant identification, v, value of optimum process
minute
for variant v, PCMins$m and number of optimum process minutes available for
Zo producing variant v, PCMins~,. The fourth column, Remaining minutes, is
empty
at this stage, but will be used later.
2. Valuing the Excess Crewing Component of a Problem p over a Base
Table Period:
Refer to Fig. 21. An important step in the valuation of the excess crewing
is component of a problem is determining XCLL,""" the total number of labor
hours
due to an excess crewing problem, p with variant v, over the base table
period.
This is done by first determining XCLL~,, the lost labor hours due to excess
crewing over a production run and then summing it over the base table period.
XCLL~, is determined by finding the difference between the optimum crew size,
so OCS" and the actual crew size ACS", and multiplying by the actual
processing
CA 02321283 2000-08-17
WO 99/42939 PCTIAU99100098
28
time of the run, APH",~, for all variants v over the production run. Step S52.
OCS" was stored in database 15 during the modeling of the process and ACS",a
and APH",d, the actual processing time of the run were acquired and entered
into
the database at the end of the production run. Having determined XCLL~,, the
total number of labor hours due to an excess crewing problem over the
production run, the total excess labor time due to a problem p for a variant
over
the base table period, XCLL~", is found by summing XCLL~ over all.shifts and
days in the base table period. Step S54. The total lost labor hours for the
whole
process XCLL~, is then found by summing XCLLp"~, for all variants in the
process.
~o Step S56.
The lost labor time due to an excess crewing problem, XCLL$~" is valued by
multiplying XCLL~, by AULR, the all up labor rate for the process (previously
stored in database 15 during the modeling stage). Step S58.
3. Valuation of Downtime, Slow Running and Bottleneck Waste
~s Problems:
A typical process has three sources of lost process time - downtime (i.e. when
the process stops entirely), slow running (i.e, when the process runs at rate
.
slower than the optimum rate) and bottleneck waste (i.e. when product is
produced at the speed bottleneck, but must be discarded, thus wasting
Zo processing time).
The manner of valuation of the lost process time depends on whether the
process is production constrained or sales constrained. A process is
production
constrained when the company is able to sell more of the product than it can
produce. It is sales constrained when the company cannot sell any more than it
2s can produce. If the process is production constrained, then some or all
lost
process time could have been used to make additional product, had the process
been operating at its optimum level. The lost process time can therefore be
valued in terms of the marginal profitability of the additional product. This
is done
by determining the value of a unit of time, assuming the process to be
operating
ao at its optimum level. If the process is sales constrained, then lost
process time is
valued in terms of labor cost savings which would result from improving the
CA 02321283 2000-08-17
WO 99/42939 PGT/AU99/00098
29
process efficiency.
In any case, the lost process time is initially valued by means of a method
called
the "Lost Production Cost Evaluation Method" also referred to as the
"Production
Constrained Valuation Method" which is a module of application 13. The Lost
s Production Cost Evaluation Method values as many of the lost process minutes
as possible in terms of marginal profitability and outputs a value of the lost
process time. Any remaining time is also output to be valued as wasted labor
time.
a. The Lost Production Cost Evaluation Method:
~o The program module which implements the Lost Production Cost Evaluation
Method uses as inputs the data in the Production Constrained Savings Table
described above, and OPM~" a quantity of lost optimum process minutes due to
problem p. The outputs of the program module are OPM$~" the maximum value
of all or part of the input lost process minutes and OPMR~" the quantity of
lost
~s process minutes which remains unvalued at the end of the algorithm.
Referring to Fig. 22, the basic steps in the Lost Production Cost Evaluation
Method are:
1. Copy the contents of the PCMins~, column of the Available Production
Constrained Savings Table into a new column called RemMins (i.e.
2o remaining minutes). Step S60.
2. Initialise the lost optimum process minute value as zero (i.e. OPM$~,= 0).
Step S62.
3. Locate the variant in the Available Production Constrained Savings Table
whose production constrained minute value, PCMins$~" is greatest and
is whose RemMins value is greater than zero. Step S64. If there is no such
variant, then the valuation is complete {i.e. the variant is sales
constrained). The program outputs OPMR~" otherwise, the program
continues. Step S66.
CA 02321283 2000-08-17
WO 99/42939 PGT/AU99/00098
4. For the variant selected in step S64, determine the lesser of the quantity
of optimum process minutes to be valued, OPM~, and RemMins. This is
called the "usable optimum process minutes," UOPM~, because these
minutes ark production constrained - they can be used to produce more
s variants which can be sold. Step S68.
5. Multiply the usable optimum process minutes, UOPM~,, by the production
constrained minute value, PCMmins$ and add the product to the lost
optimum process minute value, OPM$~,. Step S70.
6. Deduct usable optimum process minutes, UOPMp" from both the values in
~o the input lost optimum process minutes, OPM~, and the RemMins
columns. Step S72.
7. If both the optimum process minutes, OPM~, and RemMins are greater
than zero, then return to step S64. Steps S74 and S76. Otherwise the
evaluation is complete and the lost optimum process minute value,
OPM$~, and remaining optimum process minutes, OPMRp~ (i.e. those that
have not yet been valued) are output. Step S78.
b. Valuing the Process Downtime Component of a Problem p over the
BTP:
There are two components to the cost of a downtime problem. The first, and
Zo most significant, occurs where the lost time could have been utilised to
produce
more units which could have been sold. If none or only some of the units could
be sold, the remaining component of downtime loss is the cost of labor while
the
process is down. The total downtime is the sum of these two components.
Refer now to Fig. 23. The first step is the determination of how many optimum
zs process hours were lost over the BTP due to process downtime. DTI, the
downtime due to a problem p for variant v on a day and shift is retrieved from
the
database (it was entered at the end of a shift). Step S80. This is summed over
the BTP, to give DTI"" the downtime for a variant v due to problem p. Step
582.
DTOPp"", the number of optimum process hours lost due to downtime from a
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
31
problem p with variant v over the base table period is found by multiplying
DTI",
by TPE"", the true process efficiency for a particular variant over the BTP.
Step
S84.
True process efficiency is the number of optimum process hours for variant v
s over the BTP, OPH~" divided by the actual process hours over the BTP, APHm.
Step S88. APH~, is input at the data input stage. OPH~, is found by dividing
the
output of the variant over the BTP, O~" by the maximum bottleneck speed, MBS".
S86.
This process is repeated for all variants. DTOP~,, the total lost optimum
process
,o hours due to a downtime problem p is found by summing DTOP~, over all
variants. Step S90. These minutes are input to the Lost Production Cost
Evaluation Method program module and thus valued. Step S92. See Fig. 22.
The second component of downtime loss is lost labor hours. Lost labor hours
are found by summing over the BTP, the labor hours lost in each production
run.
~s The lost labor hours for a production run will vary depending on whether
the
downtime problem is also an excess crewing problem in that run. If the
downtime
problem is also an excess crewing problem, then the downtime lost labor hours,
DTLLp"",, is the product of the downtime due to problem p, DT~"~, the actual
crew
size, ACS",d and the true process efficiency, TPE~~. Step S94. If the downtime
zo problem is not an excess crew problem, then the downtime lost labor hours,
DTLL~, is the product of the downtime due to problem p, DTI, the optimum
crew size, ACS" and the true process efficiency, TPE~,. Step S96. These
values are calculated for each production run, and summed to give DTLL~"" the
total downtime labor loss for variant v over the BTP for a particular problem.
is Step S98. DTLL~", is then summed for all variants to give DTLL~" the total
downtime lost labor due to problem p. Step S100.
It will be recalled that the Lost Production Cost Evaluation Method program
module output a value of downtime optimum process minutes, DTOPM$P~ and an
amount of time which could not be valued, DTOPMR~,, due to the fact that
3o market conditions dictated that a fixed quantity of each variant could be
sold.
Those remaining process minutes are valued as lost labor hours in the
following
CA 02321283 2000-08-17
WO 99142939 PCT/AU99100098
32
manner: First, the proportion of lost labor hours which remain unvalued is
determined at Step S102, by dividing the remaining process minutes, DTOPMR~,
by the total lost process time due to downtime problem p, DTOPM~, (previously
calculated at Step S90). The proportion of unvalued downtime, DTI" is then
s multiplied by the total downtime labor loss due to problem p, DTLL~, and
then by
the all up labor rate for the process, AULR. Step S104. This gives DTLL$~" the
value of the lost labor due to downtime problem p.
The total value of a downtime problem p over the BTP, DT$~" is the sum of the
value of the lost optimum process minutes due to the problem, DTOPM$~, and
~o the value of the lost labor DTLL$~,. Step S106.
c. Valuing the Process Slow Running Component of a Problem p, over
the BTP:
As in the case of a downtime problem, a problem which causes slow running has
two possible cost components: the cost of lost production and the cost of
wasted
~s labor.
In order to value the cost of production losses due to slow running, the total
amount of optimum process time lost due to the problem during the BTP must be
found. For a single production run, the lost optimum process time due to a
slow
running problem is found by first ascertaining how many production units of
~o variant v were not produced because of the problem.
The proportional process uptime during the production run, UT~" is found by
summing the total downtime for all problems, DT~, (Step S108 in Fig. 23),
dividing
by the actual processing time in the BTP, APH~, and subtracting the result
from
1. Step S110, Fig. 23.
25 The lost optimum process time due to slow running problem p, for variant v,
is
calculated by first multiplying the proportional process uptime during the
production run UT~" by the actual processing time APH"~. The result is
multiplied by the difference between the maximum bottleneck speed of the
variant, MBS~, and the actual bottleneck speed, ABS"~, of the production run.
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99100098
33
The result is converted into lost optimum processing time, SROPM~, by dividing
the result by the maximum bottleneck speed MBS~ and then multiplying by the
true process efficiency, TPEm. Step S 120. See Fig. 24.
SROPM~ is then summed for all production runs in which the same slow
s running problem p occurred over the BTP to give SROPM~" the total lost
optimum process time due to stow running problem p over the BTP. Step S122.
In order to value SROPM~" it is fed into the Lost Production Cost Evaluation
Method program module shown in Fig. 22. Step S124. The resulting value of
lost production due to a slow running problem is SROPM$~, (i.e. output OPM$~,
~o of the Lost Production Cost Evaluation Method program module). If there is
any
remaining unvalued lost optimum process time SROPMR~, (i.e. output OPMRP~
of the Lost Production Cost Evaluation Method program module is greater than
zero), it is valued as lost labor time, in a manner which will now be
described.
The proportion of lost process time which has not been valued, SRP~" is
~s determined by dividing SROPMR~, by SROPM~" the total lost optimum process
time due to problem p over the BTP. Step S126. Next, the total quantity of
lost
labor time attributed to the slow running problem p is calculated. For each
production run, the quantity of lost optimum process hours is SROPM~, which
was calculated at Step S120. This is converted into lost labor time SRLL~,, by
multiplying it by the actual crew size, ACS~,~, for the variant being produced
if the
slow running problem is not the same as the excess crew problem on that day
and shift. Step S128. If the slow running problem is the same as the excess
crew problem on that day and shift, then lost labor time, SRLL~2, is SROPM~,
multiplied by the optimum crew size, OCS"~. Step S130. This avoids double
is counting of lost labor time. Steps S120, S128 and S130 are repeated for
each
production run in the BTP where the slow running problem occurs, and the
resulting values of SRLL~ are summed to give SRLL~" the total lost labor time
due to slow running problem p. Step S132.
The dollar value of lost labor time due to slow nrnning problem p is the
product of
3o SRP~" the proportion of optimum process hours which were not valued by the
Lost Production Cost Evaluation Method, and SRLL~" the total lost labor time
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
34
due to the slow running problem, and AULR, the all up labor rate. This value
is
called SRLL$~,. Step S134.
The total dollar value of the slow running component, SR$~" is determined by
adding the lost production component SROPM$~, and the lost labor component
s SRLL$~. Step S136.
d. Calculation of the Bottleneck Waste Component of Problem p over
the BTP:
Bottleneck waste is product which is produced at or after the speed
bottleneck,
but which must be scrapped. As in the case of downtime and slow running
~o problems, the value of bottleneck waste is made up of two components, lost
production and lost labor.
Refer to Fig. 25. Lost production is valued as follows: BWOPM~,, the amount
of optimum process time lost due to a bottleneck waste problem p, in a
particular
production run is calculated by dividing BWU~, the number of production units
~s wasted, by MBS~, the maximum bottleneck speed for the variant and
multiplying
by true process efficiency, TPEY. S140. BWU~~ is calculated by multiplying the
quantity of units of the variant rejected at or after the bottleneck
(WasteQty"~')
ascertained at the data input stage by the waste reporting unit to production
unit
factor ("WRUtoPU~;') input at the modeling stage. BWOPM~ is summed for all
2o production runs in the BTP in which the bottleneck waste problem p
occurred,
giving BWOPM~" the total amount of optimum process time lost due to
bottleneck waste problem p. Step S142.
BWOPM~, is then fed into the Lost Production Cost Evaluation Method program
module (also referred to as the "Production Constrained Valuation Box"), which
2s values as much of the lost optimum process time as possible, outputting
BWOPM$p". Step S144. The other output, BWOPMR~" the remaining lost
process time, is valued as lost labor time.
First, the proportion of lost optimum process time which has not been valued
is
calculated by dividing BWOPMR~, by BWOPM~" the total lost optimum process
CA 02321283 2000-08-17
WO 99/42939 PGT/AU99/00098
time caused by the bottleneck waste problem p, giving BWP~,. Step S146. Next,
the total lost labor hours in a single production run due to bottleneck waste
problem p is calculated. BWOPM~",~,, the amount of optimum process time lost
due to a bottleneck waste problem p, in a particular production run, is
converted
s to lost process hours, BWLL~",~", by multiplying BWOPM~ by the optimum crew
size OCS", if the bottleneck waste problem is the same as the excess crew'
problem, see Step S148. BWOPM~ is multiplied by the actual crew size,
ACS,, if the bottleneck waste problem is not the same as the excess crew
problem, giving BWLL~. Step S150. BWLL~", and BWLL~,,2 are the total
~o amount of lost labor time due to bottleneck waste problem p, in a
particular
production run or shift s and day d. BWOPM~, can also be calculated by
dividing BWU~ by MBS".
BWLL~"~, or BWLL, are summed over all production runs in the BTP, giving
the total lost labor time due to bottleneck waste problem p, BWLL~,. Step
S152.
~s This is valued by multiplying it by BWP~,, the proportion of lost labor
time due to
bottleneck waste problem p which was unvalued (see Step S146) and AULR, the
all up labor rate, resulting in BWLL$~,. Step S154. The total value of the
bottleneck waste problem p over the BTP is calculated by adding the lost
optimum production time value BWOPM$~, and BWLL$p~ the lost labor value,
Zo resulting in BW$~,. Step S156.
4. Valuing the Wasted Raw Material Component of Problem p over the
BTP:
Refer to Fig. 26. For each type of raw material wasted in the BTP due to
problem p, the total quantity of raw material wasted over the BTP, WasteQtyp~~
25 {input at the data input stage), is multiplied by the waste reporting unit
to raw
waste conversion factor {WRUtoRU~,) to give a quantity of wasted raw material,
("RWQty~""). S160. RWQty~,~ is multiplied by RW$~,~ (input at the modeling
stage), the dollar value of a single unit of the material, to give "RW$"" the
value
of that type of raw material wasted. Step S162. This is repeated for all types
of
so raw materyals used and then summed to give RW$~" the total value of raw
material wasted due to problem p in the BTP. Step S164.
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
36
5. Summing the Total Value of Problem P over the BTP:
All components of the cost of problem p have now been determined over the
BTP. The total value of the problem is found by adding all the components,
i.e.
values of excess crew, XCLL$~" downtime DT$~" slow running, SR$~"
bottleneck waste, BW$~" and raw materials, RW$~" giving PROB$~" the total
value of the problem p. Step S170 in Fig. 27.
6. Extrapolating the Cost of Problem p over a Year:
PROB$~, is now extrapolated from the BTP to an entire year by multiplying it
by
52 weeks and dividing by the number of weeks in the BTP, typically five. The
result is the total value of problem p over an entire year, PROB$p. Step S 172
in
Fig. 27. PROB$p is saved in a table in the database. Step S174.
7. Repeating Steps 1-8 for All Problems in the Process over the BTP:
The steps which have just been described are repeated for all of the problems
in
the process. Steps S176 and S178.
~s 8. Sorting the Problems by Total Value to generate the Problem Priority
Table:
The results which were stored in the table are now sorted such that the
problem
which has the highest value is at the top of the table and remaining problems
are
stored in decreasing order of value. Step S180. This is the "Problem Priority
so Table." The Problem Priority Table is then printed out. Step S182. The
plant
management can then decide which of the problems to work on and allocate
resources in accordance with the values of the problems. Step S184.
While the invention has been described with reference to its preferred
embodiment, it will be appreciated by those of ordinary skill in the art that
various
is modifications can be made to the preferred embodiment without departing
from
the spirit of the invention or limiting its scope.
CA 02321283 2000-08-17
WO 99/42939 PCT/AU99/00098
37
INDUSTRfAL APPLtCABILIT'Y
It will be appreciated by one skilled in the art that the method and computer
system of the invention have important application in maximising efficiency of
processes, especially manufacturing processes.
