U.S. patent application number 12/000341 was filed with the patent office on 2008-05-01 for project risk management system utilizing probability distributions.
Invention is credited to Kenji Araki, Makoto Kudoh, Youichi Nishikawa, Hisanori Nonaka, Takeshi Yokota.
Application Number | 20080103859 12/000341 |
Document ID | / |
Family ID | 29561445 |
Filed Date | 2008-05-01 |
United States Patent
Application |
20080103859 |
Kind Code |
A1 |
Yokota; Takeshi ; et
al. |
May 1, 2008 |
Project risk management system utilizing probability
distributions
Abstract
When the process plan made for execution is corrected based on
the process correction condition, the variation amount in each of
processes based on the process correction condition is calculated
as the probability distribution using the probability distribution
data generated by obtaining the probability distribution from the
variation-amount prediction value of each of the processes.
Thereby, the influence degree on other processes when the process
is corrected can be estimated not simply as the propagation of the
variation fixed value but so as to be more suited to the actual
circumstances in accordance with the attribute information and the
past variation patterns of the process.
Inventors: |
Yokota; Takeshi; (Hitachi,
JP) ; Nonaka; Hisanori; (Tokai, JP) ; Araki;
Kenji; (Mito, JP) ; Nishikawa; Youichi;
(Tokyo, JP) ; Kudoh; Makoto; (Kashiwa,
JP) |
Correspondence
Address: |
MATTINGLY, STANGER, MALUR & BRUNDIDGE, P.C.
1800 DIAGONAL ROAD
SUITE 370
ALEXANDRIA
VA
22314
US
|
Family ID: |
29561445 |
Appl. No.: |
12/000341 |
Filed: |
December 12, 2007 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10246690 |
Sep 19, 2002 |
7318039 |
|
|
12000341 |
Dec 12, 2007 |
|
|
|
Current U.S.
Class: |
705/7.28 |
Current CPC
Class: |
G06Q 10/0635 20130101;
G06Q 10/06313 20130101; G06Q 10/063118 20130101; G06Q 40/08
20130101 |
Class at
Publication: |
705/007 |
International
Class: |
G06Q 10/00 20060101
G06Q010/00 |
Foreign Application Data
Date |
Code |
Application Number |
May 29, 2002 |
JP |
2002-155945 |
Claims
1. A project risk management system, wherein when a process plan
made for execution is corrected based on a process correction
condition, a variation amount in each of processes based on said
process correction condition is calculated as a probability
distribution using probability distribution data generated by
obtaining a probability distribution from a variation-amount
prediction value of each of said processes.
Description
BACKGROUND OF THE INVENTION
[0001] The present invention relates to a project risk management
system for evaluating how much influence correction of a process
plan exerts on an entire project while project management is
performed.
[0002] For performing a project, when a process plan from start to
finish is worked out, each of the worked-out processes has its own
time constraint (i.e., a constraint condition determined by a
period required for completion of the process). Each of the
worked-out processes is also has a sequence constraint which is a
constraint condition determined by an execution sequence relation
with other processes. In the process plan for the project, during
execution of the process plan, a construction period might be
corrected or modified. Thus, in executing the process plan as
described above, if the construction-period correction/modification
occurs, how much influence it exerts on other processes is
evaluated.
[0003] When this construction-period correction/modification of a
certain process occurs and the influence of the
correction/modification on other processes is evaluated,
conventionally, a method is often employed in which it is
calculated on the basis of the time constraint and the sequence
constraint in each of the processes how the construction-period
correction/modification amount in the certain process propagates to
processes, which are subjected to the sequence constraint relation
with the certain process, within the range of the time constraint.
As such a conventional specific approach to managing the
correction/modification of the process plan, there is provided
JP-A-10-240804 and so forth, for example.
SUMMARY OF THE INVENTION
[0004] The conventional method of calculating how the
construction-period correction/modification amount propagates
within the range of the time constraint defines the propagation of
the correction/modification amount of the process plan as a fixed
value. That is, when execution of a process "A" is ten days behind
an initial process plan, it is calculated that a process "B"
subjected to the sequence constraint relation with the process "A"
(having the execution sequence relation with other processes) is
also delayed ten days.
[0005] However, even if the process "B" is subjected to the
sequence constraint relation with the process "A" and if the
execution of the process "A" is ten days behind the initial process
plan, countermeasures against the delay can be sometimes made, in
actuality. In such a case, the propagation of the delay amount in
the process becomes variable. Accordingly, even if the execution of
the process "A" is delayed for ten days, it often happens that the
propagation amount of the correction/modification to the other
processes does not always become ten days due to the specific
contents of the sequence constraint of the process "B" imposed on
the process "A" or other individual factors.
[0006] For this reason, when the conventional evaluation of the
influence degree of the correction/modification in the process plan
is performed, in order to obtain the more accurate propagation
amount to the other processes, there is a problem that it is
necessary to determine the influence degree in view of uncertain
factors such as the specific contents of the sequence constraint
and other individual factors.
[0007] An object of the present invention is to provide a project
risk management system and a project risk management apparatus
which can perform an evaluation of an influence degree which is
more suited to actual circumstances, upon evaluation of the
influence degree of contents of correction/modification of a
process plan on other process plans.
[0008] In order to achieve the object described above, in a project
risk management system of the present invention, an influence value
of a correction/modification amount is calculated using information
in which an influence rate of correction/modification amount of a
process is defined as a probability distribution, and information
in which an individual process correction method dependent on
various variation factors is defined as rule information, in
addition to information on a constraint condition which is
conventionally used.
[0009] Further, means for automatically generating the information
on the probability distribution and the rule information using past
instance data on project management is employed.
[0010] Other objects, features and advantages of the invention will
become apparent from the following description of the embodiments
of the invention taken in conjunction with the accompanying
drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0011] FIG. 1 is a drawing showing a basic configuration of a
project risk management system and a project risk management
apparatus according to the present invention, which is minimally
required;
[0012] FIG. 2 is a drawing showing a basic configuration of the
project risk management system and the project risk management
apparatus illustrated in FIG. 1, with system expansion functional
units included;
[0013] FIG. 3 is a drawing showing a flowchart for overall
processing by a project risk management system based on the
configuration of the project risk management system and the project
risk management apparatus illustrated in FIG. 2;
[0014] FIG. 4 is a drawing showing a data recording format for
instance data recorded in an instance database;
[0015] FIG. 5 is a drawing showing a flowchart for processing by a
probability distribution data generator;
[0016] FIG. 6 is a drawing showing a data recording format for a
probability distribution database illustrated in FIGS. 1 and 2;
[0017] FIG. 7 is a drawing showing a flowchart for processing by a
rule data generator illustrated in FIG. 2;
[0018] FIG. 8 is a drawing showing a data recording format for a
rule database illustrated in FIG. 2;
[0019] FIG. 9 is a drawing showing a display screen of a
process-correction-condition input device illustrated in FIG.
2;
[0020] FIG. 10 is a drawing showing a data recording format for a
constraint condition database illustrated in FIGS. 1 and 2;
[0021] FIG. 11 is a drawing showing a data recording format for a
process plan database illustrated in FIGS. 1 and 2;
[0022] FIG. 12 is a drawing showing a flowchart for processing by a
process correction device illustrated in FIGS. 1 and 2;
[0023] FIG. 13 is a drawing showing a display screen of process
variation amounts in the process plan display device illustrated in
FIG. 2;
[0024] FIG. 14 is a drawing showing a display screen of a frequency
distribution in the process plan display device illustrated in FIG.
2;
[0025] FIG. 15 is a drawing showing a display screen of a frequency
distribution of cost variation amounts; and
[0026] FIG. 16 is a drawing showing an example of a hardware
configuration of the project risk management system and the project
risk management apparatus according to the present invention.
DETAILED DESCRIPTION OF THE EMBODIMENTS
[0027] An embodiment of a project risk management system and a
project risk management apparatus according to the present
invention will be described with reference to FIGS. 1-16.
[0028] FIG. 1 shows a basic configuration of the project risk
management system and the project risk management apparatus
according to the present invention, which is minimally
required.
[0029] Incidentally, the description is directed to the case where
a construction work is used as an example of a project for the
project risk management system and the project risk management
apparatus according to the present embodiment.
[0030] Referring to FIG. 1, the project risk management system and
the project risk management apparatus basically include a
process-correction-condition input device 1, a constraint condition
database 2, a probability distribution database 3, a process plan
correction device 4, and a process plan database 5.
[0031] For performing correction of a process plan, it is first
inputted, using the process-correction-condition input device 1,
which process (e.g., a process "A") is to be corrected and how the
process should be corrected (e.g., ten days' delay in the process
plan). Then, processes (e.g., processes "C" and "D") associated
with the corrected process (e.g., the process "A") are extracted
using information stored in the constraint condition database 2. At
the same time, information indicating to what extent these
processes (e.g., the processes "C" and "D") can be varied is also
extracted from the constraint condition database 2.
[0032] Further, it is extracted, as information on probability
distribution, from the probability distribution database 3 how much
influence and with how much probability the processes (e.g., the
processes "C" and "D") which will be influenced are subjected to
according to the correction contents (e.g., ten day's delay in the
process plan) of the corrected process (e.g., the process "A") and
attribute information (information on the type of a work and the
type of parts to be worked on and so forth) of the processes (e.g.,
the processes "C" and "D") which will be influenced.
[0033] Using these information, for process information recorded in
the process plan database 5, the process correction device 4
calculates the influence degree on the processes (e.g., the
processes "C" and "D") which will be influenced by the process
(e.g., the process "A") to which the correction is inputted. Then,
the process correction device 4 records the result of the
calculation in the process plan database 5 as updated process
information.
[0034] FIG. 2 shows a basic configuration of the project risk
management system and the project risk management apparatus
illustrated in FIG. 1, with system expanding functions
included.
[0035] In addition to the configuration of the project risk
management system illustrated in FIG. 1, the project risk
management system and the project risk management apparatus
illustrated in FIG. 2 further includes a rule database 6, a
probability distribution data generator 7, a rule data generator 8,
an instance database 9 and a process plan display device 10.
[0036] In the configuration illustrated in FIG. 2, before
correction of the process, using information recorded in the
instance database 9 which mainly records attribute information
(information on the types of works, the types of parts to be worked
on, and so forth) of each of the processes in a past project and
information on planned and actual values for construction periods
and the number of man-hour, the project risk management system
generates information on the probability distribution of process
variation amounts in accordance with the attribute information in
the probability distribution data generator 7. The project risk
management system then records the generated information in the
probability distribution database 3. Thereafter, the project risk
management system generates a process correction rule in accordance
with the attribute information in the rule data generator 8, and
then records it in the rule database 6.
[0037] For correction of the process plan, information indicating
which process (e.g., the process "A") is to be corrected and how
the process is to be corrected (e.g., ten days' delay in the
process plan) is first inputted using the
process-correction-condition input device 1. Then, the processes
(e.g., the processes "C" and "D") associated with the corrected
process (e.g., the process "A") are extracted using the information
stored in the constraint condition database 2. At the same time,
the information indicating to what extent these processes (e.g.,
the processes "C" and "D") can be varied is also extracted from the
constraint condition database 2.
[0038] Then, it is extracted as, the information on probability
distribution, from the probability distribution database 3 how much
influence and with how much probability the processes (e.g., the
processes "C" and "D") which will be influenced are subjected to
according to the correction contents (e.g., ten day's delay in the
process plan) of the corrected process (e.g., the process "A") and
the attribute information (e.g., information on the type of a work
and the type of parts to be worked on) of the processes (e.g., the
processes "C" and "D") which will be influenced. Further, the
process correction rule corresponding to the attribute information
of the process is extracted from the rule database 6.
[0039] Using these information, the process correction device 4
calculates the influence degree on the processes (e.g., the
processes "C" and "D") which will be influenced by the process
(e.g., the process "A") to which the correction is inputted. Then,
the process correction device 4 records the calculation result in
the process plan database 5 as updated process information.
Further, the correction result of the process is displayed by the
process plan display device 10.
[0040] FIG. 3 shows a flowchart for daily process information
management as overall processing by a project risk management
system based on the configuration of the project risk management
system and the project risk management apparatus illustrated in
FIG. 2.
[0041] First, in step a01, when the correction input for every
certain period or every process is performed, the probability
distribution data on the process variation amounts is generated on
the basis of past instance data, and then recorded in the
probability distribution database 3. When the probability
distribution data is generated in step a01, a process correction
rule is periodically generated on the basis of the past instance
data, and then recorded in the rule database 6. When the rule data
is generated in step a02, it is checked in step a03 whether or not
a process correction request from a user or the system through the
process-correction-condition input device 1 is made. When it is
determined in step a03 that no process correction request from the
user or the system through the process-correction-condition input
device 1 is made (i.e., when there is no correction request), the
procedure returns to step a01.
[0042] When it is determined in step a03 that the process
correction request from the user or the system through the
process-correction-condition input device 1 is made (i.e., when the
correction request is inputted), it is inputted using the
process-correction-condition input device 1 in step a04 which
process (e.g., the process "A") is to be corrected and how the
process is to be corrected (e.g., ten days' delay in the process
plan) based on the process correction request. When the process
correction condition is inputted to the system based on the process
correction request in step a04, the influence of the correction
contents of the corrected process (e.g., the process "A") on other
processes is calculated in step a05 using information on the
constraint condition data extracted from the constraint condition
database 2 based on the inputted process correction condition
(e.g., ten days' delay in the process plan), the probability
distribution data extracted from the probability distribution
database 3 based on the information on the processes (e.g., the
processes "C" and "D") associated with the corrected process (e.g.,
the process "A") and the rule data extracted from the rule database
6.
[0043] When the influence of the correction contents of the
corrected process on other processes is calculated in step a05
using information on the constraint condition data extracted on the
basis of the inputted process correction condition, the probability
distribution data extracted on the basis of the information on the
processes associated with the corrected process and the rule data,
the process plan corrected in step a05 is recorded in the process
plan database 5 in step a06. When a request from the user or the
system is made, the corrected process plan is displayed.
[0044] The embodiment of the project risk management system and the
project risk management apparatus according to the present
invention will be described below in further detail, using specific
examples.
[0045] FIG. 4 shows a specific example of the instance data
recorded in the instance database 9. Basically, items shown in FIG.
4 for respective detailed processes of a project are recorded as a
database.
[0046] Referring to FIG. 4, specifically, the items to be recorded
include a process name 901, a work type 902 of a work to be
performed in the process, a type 903 of parts to be handled in the
process, a unit number 904 of the parts to be handled in the
process, a total weight 905 of the parts to be handled in the
process, an occupational category 906 of workers who mainly perform
the work to be executed in the process, a company name 907 of a
company which mainly performs the work in the process, a planned
number of man-hour 908 for the process, an actual number of
man-hour 909 for the process, a planned construction period 910 of
the process, an actual construction period 911 of the process,
weather conditions 912 during a period in which the process is
performed, a work area 913 of the process, information 914
indicating whether or not scaffolding is used in the work of the
process, information 915 indicating whether or not a
material-handling device is used in the work of the process, and a
number 916 of workers supplied to the process.
[0047] FIG. 5 shows a flowchart for processing by the probability
distribution data generator 7.
[0048] First, the information on past instances recorded in the
instance database 9 is classified into groups in matrix [(TYPES OF
WORKS).times.(TYPES OF PARTS TO BE WORKED ON)] according to the
types of works and the types of parts to be worked on. When the
information on the past instances is classified into the groups
according to the types of the works and the types of the parts to
be worked on in step 701, one type of the works classified in step
701 is selected in step 702. When one type of the works is selected
from the types of the works classified in step 702, one type of the
parts to be worked on is selected in step 703 from the types of the
parts classified in step 701.
[0049] When one type of the works is selected in step 702 from the
types of the works classified in step 701 and one type of the parts
to be worked on is selected in step 703 from the types of the parts
classified in step 701, one process is selected in step 704 from
processes which belong to a taxonomic group defined by the type of
the work selected in step 702 and the type of the parts to be
worked on selected in step 703. When one process is selected from
the processes which belong to the taxonomic group in step 704, a
deviation amount M of the number of man-hour per unit number of the
parts for the process selected in step 704 is determined from
Equation (1), in step 705. M=((ACTUAL NUMBER OF MAN-HOUR)-(PLANNED
NUMBER OF MAN-HOUR))/(NUMBER OF PARTS) (1)
[0050] When the deviation amount M of the number of man-hour per
unit number of parts is determined in step 705, the deviation
amount M of the number of man-hour per unit number of parts for the
selected process is added to information on the value of the
deviation amount M of the number of man-hour per unit number of
parts for other processes in the selected group. On the basis of
these deviation values, frequency distribution information
(probability distribution data) on the deviation amount M of the
number of man-hour is generated as the probability distribution
data on the number of man-hour for the selected group. As segment
information on the frequency distribution, data obtained by
dividing a range from -50 to 50 days by the unit of 0.5 day is
used. When the probability distribution data on the deviation
amount M of the number of man-hour is generated in step 706, a
deviation amount T of the construction period per unit number of
parts for the process selected in step 704 is determined from
Equation (2). T=((ACTUAL CONSTRUCTION PERIOD)-(PLANNED CONSTRUCTION
PERIOD))/(NUMBER OF PARTS) (2)
[0051] When the deviation amount T of the construction period per
unit number of parts is determined in step 707, the deviation
amount T of the construction period per unit number of parts for
the selected process is added to information on the value of the
deviations amount T of the construction period per unit number of
parts for other processes in the selected group. On the basis of
these values, frequency distribution information on the deviation
amount T of the construction period (probability distribution data)
on the construction period is generated as the probability
distribution data on the construction period in the selected group.
As segment information on the frequency distribution, data obtained
by dividing a range from -50 to 50 days by the unit of 0.5 day is
used.
[0052] When the probability distribution data on the deviation
amount T of the construction period is generated in step 708, it is
determined in step 709 whether or not the processing is performed
on all the process information in the selected group. When it is
determined in step 709 that the processing is not performed on all
the process information in the selected group (or when the
processing on all the process information is not completed), the
procedure returns to step 704. On the other hand, when it is
determined in step 709 that the processing is performed on all the
process information in the selected group (or when the processing
on all the process information is completed), it is determined in
step 710 whether or not the processing is performed on all the
types of the parts to be worked on in the selected type of the
work. When it is determined in step 710 that the processing is not
been performed on all the types of the parts to be worked on in the
selected type of the work, the procedure returns to step 703. On
the other hand, when it is determined in step 710 that the
processing is performed on all the types of the parts to be worked
on in the selected type of the work (or when the processing on all
the types of the parts to be worked on in the selected type of the
work is completed), it is determined in step 711 whether or not the
processing on all the types of the works is performed. When it is
determined in step 711 that the processing on all the types of the
works is not performed, the procedure returns to step 702.
[0053] On the other hand, when it is determined in step 711 that
the processing on all the types of the works is performed, the
generated probability distribution data is outputted in step 712.
Then, the generated probability distribution data is recorded in
the probability distribution database 3 every element of the
matrix, which comprises the types of works and the types of parts
to be worked on in the works, for the number of man-hour and the
construction period, thereby finishing this processing flow.
[0054] FIG. 6 shows a data recording format for the probability
distribution database 3 illustrated in FIGS. 1 and 2.
[0055] Referring to FIG. 6, the probability distribution database 3
takes on a three-dimensional database structure which comprises
three axes including a work type 301, a type 302 of parts to be
worked on in the work, a frequency distribution segment 303
obtained by dividing the range from -50 days to +50 days by the
unit of 0.5 days. Then, the information on the frequency
distribution of the number of man-hour (M) and the information on
the frequency distribution of the construction periods (T) are
recorded in the respective elements of the probability distribution
database 3 illustrated in FIG. 6 as the frequency information for
the respective elements (the information in the respective segments
among the information on the frequency distribution of a group
defined by the type of works and the type of parts to be worked on
in the works).
[0056] FIG. 7 shows a flowchart for processing by the rule data
generator 8 illustrated in FIG. 2.
[0057] First, in step 801, one process is selected from all the
process information recorded in the instance database 9 illustrated
in FIG. 2. When one process is selected from all the process
information in step 801, the deviation amount M of the number of
man-hour per number of parts for the process selected in step 801,
which is defined by Equation (1), is calculated in step 802. When
the deviation amount M of the number of man-hour per number of
parts is calculated in step 802, the deviation amount T of the
construction period per number of parts for the process selected in
step 801, which is defined by Equation (2), is calculated in step
803. When the deviation amount T of the construction period per
number of parts is calculated in step 803, the characteristic
quantity of the process selected in step 801 is set in the form of
a quantitative value. Specifically, information (such as a fine-day
rate) recorded as the quantitative value is used without
alteration, while information (such as presence or absence of
scaffolding) recorded as a qualitative value is coded (example:
(presence of scaffolding)=1, (absence of scaffolding)=0, and so
forth), and then the coded values are used.
[0058] When the characteristic quantities (the quantitative values)
are set in step 804, the deviation amount M of the number of
man-hour per number of parts calculated in step 802 and the
deviation amount T of the construction period per number of parts
calculated in step 803 are associated (or paired) with the process
characteristic quantity (the quantitative value) set in step 804,
and recorded in the vector form (one-dimensional vector) in step
805. For example, in the forgoing example, they are recorded as the
one-dimensional vector (a code indicating the deviation amounts M
and T, fine-day rate, and the presence or absence of scaffolding).
When the deviation amount M of the number of man-hour per number of
parts is paired with the deviation amount T of the construction
period per number of parts, and then they are recorded in the
vector form in step 805, it is determined in step 806 whether or
not the processing from step 802 to step 805 is performed on all
the processes recorded in the instance database 9. When it is
determined in step 806 that the processing from step 802 to step
805 is not performed on all the processes recorded in the instance
database 9, the procedure returns to step 801.
[0059] On the other hand, when it is determined in step 806 that
the processing from step 802 to step 805 is performed on all the
processes recorded in the instance database 9 (or all the processes
are subjected to processing), one of the vectors recorded in step
805 is selected in step 807. When one vector is selected in step
807 from the vectors recorded, the vector selected in step 807 is
normalized into a unit vector in step 808. After normalization of
the vector selected in step 807 into the unit vector, angles made
between the vector normalized in step 808 and all the other vectors
(already normalized) are calculated (or the inner products of the
unit vector with all the other vectors are calculated) in step
809.
[0060] After the angles made between the vector normalized in step
808 and all the other vectors (already normalized) are calculated
(or the inner products of the unit vector with all the other
vectors are calculated) in step 809, it is determined in step 810
whether or not the number of vectors with the values of the angles
calculated in step 809 equal to or less than a given value (e.g.,
less than 10.degree.) is equal to or more than a specified value
(e.g., equal to or more than 5% of the number of all the vectors).
When it is determined in step 810 that the number of vectors with
the values of the angles calculated in step 809 equal to or less
than the given value (e.g., less than 10.degree.) is equal to or
more than the specified value (e.g. equal to or more than 5% of the
number of all the vectors), a rule is created based on a
combination of vector elements in step 811. Specifically, if the
average of the vectors in which the number of vectors of the same
direction is equal to or more than the specified number) is (M=1.5
(man-hour), T=2 (days), fine-day rate=70%, and presence or absence
of scaffolding=present), rule "If the fine-day rate is equal to or
more than 70% and scaffolding is present, 1.5 man-hour should be
incremented to the number of man-hour, and 2 days should be
incremented to the construction period) is created, for
example.
[0061] On the other hand, when it is determined in step 810 that
the number of vectors with the values of the angles calculated in
step 809 equal to or less than the given value (e.g., less than
10.degree.) is not equal to or more than the specified value (e.g.,
equal to or more than 5% of the number of all the vectors), or when
the rule is created in step 811 based on the combination of vector
elements, it is determined in step 812 whether or not the
processing is performed on all the vectors recorded in step 805.
When it is determined in step 812 that the processing is not
performed on all the vectors recorded in step 805, the procedure
returns to step 807. On the other hand, when it is determined in
step 812 that the processing is performed on all the vectors
recorded in step 805 (or all the vectors are subjected to the
processing), the rule created in step 811 is outputted to the rule
database 6 in step 813, thereby completing this processing flow. If
a plurality of rules with the deviation amounts of quantitative
value parameters being within .+-.10% are present in this
processing flow, a rule created using the average of the
quantitative values for these rules is recorded as a typical
rule.
[0062] FIG. 8 shows a data recording format for the rule database 6
illustrated in FIG. 2.
[0063] Referring to FIG. 8, data in the rule database 6 is
classified into data in a condition division and data in a rule
division, and is recorded for each rule. In the condition division
of the rule database 6, parameters 601 used as criteria,
application conditions 602 of these parameters, specific thresholds
603 for the application conditions, and relations 604 with other
condition when a plurality of the conditions are present are
recorded. Then, as a rule in the rule database 6 to be executed if
the condition is satisfied, a parameter 605 which will be
influenced by execution of the rule and an influence degree 606 are
recorded. For example, rule data in the data recording format for
the rule database 6 in FIG. 8 indicates that "if the fine-day rate
is equal to or more than 70% and scaffolding is present, the
construction period is incremented by 2 days".
[0064] FIG. 9 shows a display screen of the
process-correction-condition input device 1 illustrated in FIG.
2.
[0065] Referring to FIG. 9, it is inputted using the display screen
of the process-correction-condition input device 1 which process is
to be corrected and to what extent the process is to be corrected.
On the display screen illustrated in FIG. 9, the processes recorded
in the process plan database 5 are displayed as a process table
101, so that the user selects on the display screen of the process
table 101 a process which he desires to correct, such as a process
102, by means of a mouse and so forth. At this point, attribute
information 104 of the process 102 and information 105 indicating
the execution period of the process 102 and the number of man-hour
for the process 102 are displayed as information on the selected
process 102. Then, the user sets how to correct the selected
process 102, by setting the execution period of the process after
correction as a process 103 on the process table, or by inputting
from an input interface 106 information on the start and finish
dates of the process and the number of man-hour for the process as
numerical values. Incidentally, if the process correction degree is
specified on the process table, the process 103 automatically
causes the contents of correction to be reflected on the display
contents of the input interface 106. If the specification is made
using the input interface 106, the contents of correction is
reflected as the process 103 on the display of the process
table.
[0066] Further, the number of times of a Monte Carlo simulation
which is performed when the deviation amount of the process is
estimated is set using an entry box 107. The Monte Carlo simulation
is a mathematical approach to determine the approximate solution of
a problem by numerical experiments using repetitive calculations
and random numbers, and simulates the variation amount of the
process using random numbers for stochastic variations by means of
a computer. Furthermore, a percentile value for evaluating the
variation-amount prediction values calculated as the probability
distribution is set using an entry box 109. The percentile value is
a parameter used for presenting the variation amount which occurs
with a probability equal to or below that percentile. Then, by
selecting a "correction execution" button 108 after completion of
all the settings, an estimation process of the variation amount is
started under the set conditions.
[0067] FIG. 10 shows a data recording format for the constraint
condition database 2 illustrated in FIGS. 1 and 2.
[0068] Referring to FIG. 10, as the constraint data in the rule
database 6, a process code 201 (a process 406) for uniquely
defining a process (or identifying a process), information on a
time constraint (the constraint condition determined by a period
required for completion of a process), and information on a
sequence constraint (the constraint condition determined by the
execution sequence relation with other processes) are recorded for
each process. The time constraint illustrated in FIG. 10 which is
data of the constraint condition for the rule database 6 defines
the period over which the process can be performed, and is defined
by a start time range 202 (e.g., from Aug. 1, 2001 to Oct. 31,
2001) and an finish time range 203 (e.g., from Sep. 20, 2001 to
Dec. 1, 2001). When correcting a process, the process is corrected
within the start time range and the finish time range set by the
time constraint.
[0069] On the other hand, the sequence constraint illustrated in
FIG. 10 which is data of the constraint condition for the rule
database 6 defines processes subjected to the execution sequence
with respect to the pertinent process, with a process code. In a
preceding process code 204, processes (processes 101, 102 and 205)
which is to be finished before the start of the pertinent process
are defined. In a subsequent process code 205, processes (processes
412 and 501) which cannot be started unless the pertinent process
is finished) are defined.
[0070] FIG. 11 shows a data recording format for the process plan
database 5 illustrated in FIGS. 1 and 2.
[0071] Referring to FIG. 11, items shown in FIG. 11 for each
detailed process are recorded in the process plan database 5 as a
database. Specifically, the items to be recorded includes a process
name 501 indicating the name of the process, a process code 502 for
uniquely defining the process, a work type 503 indicating the type
of a work to be performed in the process, a type of parts to be
handled in the process 504, a unit number of the parts to be
handled in the process 505, a total weight of the parts to be
handled in the process 506, an occupational category of workers 507
who mainly perform the work to be executed in the process, a
company name indicating the name of a company which mainly performs
the work in the process 508, a planned number of man-hour 509 at
the planning stage of the process, a planned construction period
510 in the planning stage of the process, a planned start date 511
of the process, a planned finish date 512 of the process, a work
area 513 of the process, information 514 indicating whether or not
the work of the process requires scaffolding, information 515
indicating whether or not the work of the process requires a
material-handling device, a planned number 516 of workers to be
supplied to the process, data 517 of frequency distribution on
variation-range prediction values for the start date of the
process, data 518 of frequency distribution data on variation-range
prediction values for the finish date of the process, data 519 of
frequency distribution data on variation-amount prediction values
for the construction period, and data 520 of frequency distribution
on variation-amount prediction values for the number of man-hour.
For the data of the frequency distributions of the respective
prediction values, the values are not recorded before processing by
the process correction device 4 is not performed.
[0072] FIG. 12 shows a flowchart for processing by the process
correction device 4 illustrated in FIGS. 1 and 2.
[0073] First, in step 401, a parameter S for a counter indicating a
simulation count displayed on the display screen of the
process-correction-condition input device 1 illustrated in FIG. 2
is set to "1". When the parameter S indicating the simulation count
is set to "1" in step 401, processes are extracted which are
subjected to the sequence constraint with respect to a process for
which the input for correction is performed by the user or the
system. When the processes subjected to the sequence constraint
with respect to the corrected process are extracted in step 402,
one of the processes extracted in step 402 is selected in step
403.
[0074] When one of the processes extracted in step 402 is selected
in step 403, the variation-amount prediction value of the process
selected in step 403 is calculated as the sum of a variation amount
in the corrected process, a variation amount randomly selected on
the basis of the probability distribution data, and a variation
amount defined by rule data the condition of which is satisfied by
the attribute data of the process selected in step 403. (PROCESS
VARIATION AMOUNT)=(VARIATION AMOUNT IN CORRECTED
PROCESS)+(VARIATION AMOUNT DETERMINED BY PROBABILITY DISTRIBUTION
DATA)+(VARIATION AMOUNT DETERMINED BY RULE DATA) For calculation,
the variation amount in the construction period is used as the
variation amount without alteration. However, the variation amount
in the number of man-hour is converted to the value expressed in
terms of the variation amount in the construction period, based on
the number of workers and using Equation (3). (CONSTRUCTION PERIOD
VARIATION AMOUNT)=(VARIATION AMOUNT IN NUMBER OF MAN-HOUR)/(NUMBER
OF WORKERS) (3)
[0075] Then, using information on the prediction value for the
variation amount in the construction period, the prediction values
for the variation ranges of the start and finish dates of the
process are calculated.
[0076] When the variation amount prediction values for the process
selected in step 403 are calculated in step 404, it is determined
in step 405 whether or not the processing is performed on all the
processes subjected to the sequence constraint, which are extracted
in step 402. When it is determined in step 405 that the processing
is not performed on all the processes subjected to the sequence
constraint, which are extracted in step 402, the procedure returns
to step 403. On the other hand, when it is determined in step 405
that the processing is performed on all the processes subjected to
the sequence constraint, which are extracted in step 402, the
processing from step 402 to step 406 is performed on the processes
subjected to the sequence constraint with respect to all the
processes extracted in step 402.
[0077] When the processing from step 402 to step 406 is performed
on the processes subjected to the sequence constraints with respect
to all the processes extracted in step 402, the variation amounts
in the respective processes corrected by a series of processing are
handled as frequency data in the current simulation count, so that
the information on the variation-amount frequency distribution data
for the respective processes is updated. When the variation amounts
in the respective processes corrected by a series of processing are
handled as the frequency data in the current simulation count and
then the information on the variation-amount frequency distribution
data for the respective processes is updated in step 407, the
parameter S is incremented by "1" in step 408.
[0078] When the incrementing process is performed in step 408, it
is determined in step 409 whether or not the parameter S exceeds
the simulation count set by the process-correction-condition input
device 1. When it is determined in step 409 that the parameter S
does not exceed the simulation count set by the
process-correction-condition input device 1, the procedure returns
to step 403. On the other hand, when it is determined in step 409
that the parameter S exceeds the simulation count set by the
process-correction-condition input device 1, the variation-amount
frequency distribution data on the process plan after correction is
outputted in step 410. Thereafter, the variation-amount frequency
distribution data on the respective processes is recorded in the
process plan database 5, and then this processing flow is
finished.
[0079] FIG. 13 shows a display screen of the process plan display
device 10 illustrated in FIG. 2, which indicates the process
variation amounts.
[0080] Referring to FIG. 13, a process table 1001 is displayed on
the display screen of the process plan display device 10 indicating
the process variation amounts. This process table 1001 displays
information on a correction target process 1002 set by means of the
process-correction-condition input device 1 illustrated in FIG. 9,
a process 1003 resulting from correction of the correction target
process 1002, and ranges of variation amounts (variation ranges
occurring with a probability set in terms of a percentile value)
1009 of respective processes. On the display screen of the process
variation amounts in the process plan display device 10 illustrated
in FIG. 13, a process 1008 in which the process selected by the
user by means of a mouse and so forth is displayed in a different
color, and attribute information 1004, construction period
information 1005 and construction period variation amount
information 1006 of the user-selected process 1008 are displayed.
Further, by specifying (or clicking) a "detailed distribution"
button 1010 on the display screen of the process plan display
device 10 illustrated in FIG. 13, indicating the process variation
amounts, a shift to a display for displaying information on the
variation-amount frequency distribution of the user-selected
process 1008 takes place. The display screen of the process plan
display device 10 illustrated in FIG. 13, indicating the process
variation amounts also displays information 1007 on the simulation
count and percentile information 1011 set by the
process-correction-condition input device 1.
[0081] FIG. 14 shows a display screen of the process plan display
device 10 illustrated in FIG. 2, indicating the frequency
distribution.
[0082] Referring to FIG. 14, first, a selection switch 1013 is used
to select which information of the frequency distribution to be
displayed for a selected process. At the time of display, four
items "start date", "finish date", "construction period" and
"number of man-hour" become the display targets. Then, a frequency
distribution graph 1012 concerning the item set by the selection
switch 1013 is displayed. To take the frequency distribution of the
start dates as an example (refer to the selection switch 1013), the
horizontal axis of the frequency distribution graph represents the
date, while the vertical axis of the graph represents the
frequency. Further, the planned date at the planning time of the
start date of the selected process is indicated by a dotted line
1014, and the variation range 1015 of the start date based on the
set percentile is also displayed. The attribute information 1016,
construction period information 1018 and percentile information
1017 of the selected process are also displayed. Further, by
pressing (or clicking) an "entire display" button 1019, the
probability distribution of the construction period variation
amount of not each process but the overall project can also be
displayed. Incidentally, the method of displaying the probability
distribution for this purpose is the same as the method of
displaying the probability distribution for each process.
[0083] In this embodiment, basically, the variation amount in the
construction period is predicted. However, by employing a model (in
a linear form and so forth) which defines the relationship between
the variation amount in the construction period and the resulting
increased cost, it also becomes possible to evaluate the influence
of the process correction on future cost variations. The display
screen showing the cost variation amount in that case is shown in
FIG. 15.
[0084] Referring to FIG. 15, a cost frequency distribution graph
1020 is displayed on the display screen. The horizontal axis of the
graph represents the cost amount, while the vertical axis of the
graph represents the frequency. On the display screen illustrated
in FIG. 15, a planned cost at the planning time is indicated by a
dotted line 1021, and a cost variation range 1022 based on the set
percentile value is also displayed. On a display section 1023 of
the display screen illustrated in FIG. 15, the planned cost, an
expected cost value and a cost percentile are displayed as specific
amounts of money. On the display section 1023 of the display screen
illustrated in FIG. 15, a percentile value 1024 is also
displayed.
[0085] In this embodiment, the description is directed to the
instance of the project management in the construction work.
However, the present invention is not limited to the project
management in the construction work, and can also be extensively
applied to the process management in other fields such as software
development, operation schedules of transportation means and a
semiconductor manufacturing process. Thus, the approach which is
the same as the one according to the embodiment can be adopted.
[0086] The foregoing description is given about an example of the
processing according to the present invention. When the processing
is actually systematized, as shown in FIG. 16, in hardware which
includes a CPU b01, a memory b02, a storage device b03, an input
device b04 and an output device b05, programs and databases
according to the present invention recorded in the storage device
b03 are transferred to the memory b02, and the processing is
performed by the CPU b01 on the basis of command information
supplied from the input device b04. Then, the result of the
processing is displayed on the output device b05.
[0087] According to the present invention, the influence amount to
other processes when a process is corrected can be estimated not as
the propagation of the variation fixed values, but so as to be more
suited to the actual circumstances according to the attribute
information of the process or the past variation patterns of the
process.
[0088] It should be further understood by those skilled in the art
that although the foregoing description has been made on
embodiments of the invention, the invention is not limited thereto
and various changes and modifications may be made without departing
from the spirit of the invention and the scope of the appended
claims.
* * * * *