JP2003030395A - Method, device, program, and recording medium for project management - Google Patents

Abstract

PROBLEM TO BE SOLVED: To quantize a risk of backing and parallel stages in project management. SOLUTION: A design structure matrix representing the degree of dependency showing dependency among a plurality of tasks included in a project in matrix form is generated first. Then the completion time of each task based upon the start time of the project is computed according to the operation period and precedence relation of the plurality of tasks. The completion time of each task is recorded at a diagonal part of the design structure matrix. Then the backing period from the completion time of a 1st task among the plurality of tasks to the start time of a 2nd task precedent to the 1st task is computed. Lastly, the degree of the dependency of the 1st task on the 2nd task is multiplied by the backing period to computes the degree of backing influence.

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to a task / risk management support technique for a product development project or the like.

[0002]

2. Description of the Related Art In a conventional project management system, an administrator first inputs a process required to carry out a project and sets a scheduled start date / finished date for each process to set a project schedule. The progress was managed by making a plan and inputting the actual start date / actual end date. Displaying each process as a bar chart from the start point to the end point is generally called a Gantt chart, which is a basic tool for project scheduling and progress management.

In project management based on the Gantt chart, it is difficult to explicitly realize a dependency relationship between tasks and a backtracking relationship. Therefore, DSM (Design Structure Matrix) was proposed as a new project management method. DSM expresses the tasks included in a project in a matrix format in which both rows and columns are arranged, with the tasks arranged in the row (or column) direction as the sender and the tasks arranged in the column (or row) direction as the receiver. The project management method is characterized by expressing the interrelationships such as the input / output relationships between the tasks in an easy-to-understand manner by writing the dependency relationships between the respective tasks in the corresponding portions of the matrix.

The DSM is suitable for simply expressing the dependency relationships between tasks of a complex process and analyzing task relevance. For more information on DSM, see MIT D
You can refer to it on the SM homepage (http://web.mit.edu/dsm/).

Conventional FMEA (Failure mode & Effect)
In the project risk management method based on s Analysis), the manager analyzes and organizes the risks of the product or process, and the result is the failure mode, the cause of the failure, and the effect of the failure (local effect, end effect). Impact)
, Failure detection method, and RPN indicating the magnitude of risk
(Risk priority number) and other items such as countermeasures against risks are described in the FMEA table to manage risk.

Generally, risk management is composed of four steps of "identification of risk", "quantification of risk", "preparation of countermeasures" and "control of countermeasures". FM
The EA supports these basic steps in risk management,
The FMEA table, which is the result of analysis by FMEA, provides a standard document for risk management.

[0007]

One of the major problems in the conventional project management system has been the inability to explicitly express the backtracking process that frequently occurs in an actual design project.

In order to solve this, a project management method by DSM has been proposed, but the present DSM framework does not include a means for quantifying the magnitude of the risk of a backtracking process. In order to evaluate the risk of the backtracking process, it is necessary to expand the project management method by DSM and establish a means to quantify the magnitude of the risk of the backtracking process.

Another problem with current DSM project management systems is that they do not include a means to quantify the magnitude of risk due to parallelization of tasks.

In an actual design project, succeeding tasks are started in parallel before the completion of a preceding task having a dependency relationship to shorten the development period. In this case, there is a risk that some problem is discovered when the preceding task is completed, the succeeding task is affected, and the work already performed by the succeeding task is wasted. It is necessary to build a means to quantify the risk of parallelizing such tasks.

Further, in the conventional project risk management method by FMEA, in order to quantify the risk, a numerical value called RPN (risk priority number), which is a product of the frequency of occurrence, the degree of influence and the degree of detection, is used. Used.

Although the frequency of occurrence, the degree of influence, and the degree of detection are determined based on pre-defined criteria, it is generally difficult and purely objective to determine the RPN in the current FMEA framework, and many In this case, the RPN value is subjectively determined based on the experience of the analyst.

In order to objectively calculate the RPN value, a systematic risk quantification method linked with a process model is required. In particular, many risks are closely related to the backtracking process, and quantifying the risk of the backtracking process is
is important. It is also important to quantify the risks due to parallelization of tasks.

An object of the present invention is to accurately perform risk analysis by FMEA based on a detailed process model represented by DSM data, and to quantify the risk of backtracking process and parallel process. .

[0015]

[Means for Solving the Problems] (1) A project management method according to the present invention creates a design structure matrix in which a degree of dependency indicating a dependency between a plurality of tasks included in a project is expressed in a matrix format. A first step of calculating a completion time of each task based on a start time of the project based on a work period of the plurality of tasks and a preceding relationship; and a first step of the plurality of tasks. The third step of calculating a backtracking period from the completion time of the task to the start time of the second task preceding the first task, the dependency from the first task to the second task, and the backtracking period. And a fourth step of calculating the backward influence degree by multiplication.

(2) The project management method of the present invention is
Based on the first step of creating a design structure matrix showing the degree of dependency indicating the dependency among a plurality of tasks included in the project in a matrix format, and the work period and the preceding relationship of the plurality of tasks, A second step of calculating the completion time of each task based on the start time of the project; a first task of the plurality of tasks; and a dependency on the first task, and the first task And a second task having a parallel relationship that starts before the completion of the first task, a third step of calculating a parallel period in which the first task and the second task overlap, and a dependency degree from the first task to the second task. And a fourth step of calculating a parallel influence degree by multiplying the parallel period.

The project management method of the present invention further comprises a fifth step of calculating a backtracking period from the completion time of the second task to the start time of the first task, and from the second task to the first task. The sixth step of calculating the backward influence degree of the first and second tasks having the parallel relationship by multiplying the dependency degree of 1 and the backward return period.

(3) In the above method (1) or (2), the completion times of the plurality of tasks are recorded in the diagonal line portion of the design structure matrix.

Of the completion times of the plurality of tasks, the completion times of the tasks forming the critical path are recorded separately from the completion times of the other tasks.

The plurality of tasks are arranged in the design structure matrix in ascending order of start time, and the triangular matrix portion on one side of the diagonal line portion has a preceding relationship and the triangular matrix portion on the other side moves backward. It represents a relationship.

(4) The program of the present invention is based on the above (1)
Alternatively, the first to fourth steps of (2) are executed. Also,
The recording medium of the present invention records a program for executing the first to fourth steps of (1) or (2) described above.

(5) The project management device of the present invention is
A design structure matrix in which the degree of dependency indicating a dependency between a plurality of tasks included in a project is expressed in a matrix format, a database in which work periods and precedence relationships of the plurality of tasks are recorded, and the work period and the precedence. Based on the relationship, with a computing device for calculating the completion time of each task based on the start time of the project, the computing device, among the plurality of tasks,
A backtrack period from the completion time of the first task to the start time of a second task preceding the first task is calculated, and the dependency degree from the first task to the second task and the backtrack period are calculated. Multiplication is performed to calculate the backtracking impact.

(6) The project management device of the present invention is
A design structure matrix in which the degree of dependency indicating a dependency between a plurality of tasks included in a project is expressed in a matrix format, a database in which work periods and precedence relationships of the plurality of tasks are recorded, and the work period and the precedence. Based on the relationship, with a computing device for calculating the completion time of each task based on the start time of the project, the computing device, among the plurality of tasks,
A first task and a degree of dependence on the first task,
Calculating a parallel period in which the first and second tasks overlap with respect to a second task having a parallel relationship that starts before the first task completes, and the dependence of the first task on the second task The degree of parallel influence is calculated by multiplying the degree and the parallel period.

The arithmetic unit further calculates a backtracking period from the completion time of the second task to the start time of the first task, and determines the degree of dependence of the second task on the first task and the dependency. The regressive period is multiplied to calculate the regressive influence degree of the first and second tasks having the parallel relationship.

(7) In the above apparatus (5) or (6), the completion times of the plurality of tasks are recorded in the diagonal line portion of the design structure matrix.

Of the completion times of the plurality of tasks, the completion times of the tasks forming the critical path are recorded separately from the completion times of the other tasks.

The plurality of tasks are arranged in the design structure matrix in ascending order of start time, and the triangular matrix portion on one side of the diagonal line portion has a preceding relationship and the triangular matrix portion on the other side moves backward. It represents a relationship.

(8) The design structure matrix of the present invention represents the degree of dependency indicating the dependency among a plurality of tasks included in a project in a matrix format, and the plurality of tasks have a short start time. In order, as an input task in the row direction and as an output task in the column direction, they are arranged in the design structure matrix, respectively, and the completion time of each task based on the start time of the project is the design structure matrix.
The triangular matrix portion on the input task side is arranged in the diagonal line portion of the matrix, and the triangular matrix portion on the input task side shows the backward relationship with respect to the diagonal line portion, and the triangular matrix portion on the output task side shows the preceding relation.

Of the completion times of the plurality of tasks, the completion times of the tasks forming the critical path are recorded separately from the completion times of the other tasks.

The plurality of tasks as the output tasks are divided into bands according to the preceding relationship of the plurality of tasks, and the work period of the plurality of tasks is the time of the preceding task preceding the completion time of the succeeding task. It is equal to the period less the completion time.

(9) In the present invention, first, the completion time and the critical path of each task are calculated based on the work periods and the precedence relations of the tasks included in the project. Next, by multiplying the dependency of the backtracking process between the two tasks and the difference between the completion times of the two tasks and the work time of the preceding task (backtracking period), the backtracking impact degree is calculated. calculate. This backtracking influence degree is an index that quantifies the risk of the backtracking process.

Further, according to the present invention, first, the completion time and the critical path of each task are calculated based on the work periods and the preceding relationships of the tasks included in the project. Next, in the parallel process, the overlapping period (parallel period) of the preceding task and the subsequent task in the parallel process is calculated based on the time when the preceding task is completed, the time when the succeeding task is completed, and the work period of the succeeding task. And
The degree of parallel influence of the parallel process is calculated by multiplying the parallel period and the dependency of the preceding task and the succeeding task in the parallel process. This degree of parallel influence is an index that quantifies the risk of parallel processes.

[0033]

BEST MODE FOR CARRYING OUT THE INVENTION A project management method, apparatus, program and recording medium according to the present invention will be described in detail below with reference to the drawings.

[Outline of Apparatus] FIG. 1 shows a schematic view of an apparatus for executing the project management method according to the present invention.

The device 1 according to this example comprises a display device 2, a database 3, an arithmetic unit 4, and an interface 5. The database 3 includes a list of tasks included in the project and D representing the dependency relationships between the tasks.
SM data and work period information of each task are accumulated. The arithmetic unit 4 refers to the DSM data and the task work period information to calculate the critical path of the project and the return-back influence degree.

The DSM data, task work period information, the project critical path, and the backtracking influence degree are displayed on the display device 2, respectively. When adding / deleting / correcting the work period information of DSM data and tasks, the work period information of DSM data and tasks accumulated in the database 3 is added / deleted / corrected via the interface 5.

[DSM] FIG. 2 shows an example of DSM data in which dependency relationships between tasks are expressed in a matrix format.

For the sake of simplicity, the tasks related to the general product development process will be described here.

The first line of DSM is the input task section 6, DS
The first column of M is the output task section 7, and the region of the second row and second column of the DSM is a dependency relationship expression section 8 that represents the dependency relationship between the input task and the output task.

The same tasks are arranged in the same order in the input task section 6 and the output task section 7. The numbers in the dependency relation expression unit 8 indicate that there is a dependency relation between the input task and the output task, and the size of the numbers indicates the dependency degree between the tasks multiplied by 10. It is assumed that the larger the number, the higher the degree of dependency, that is, the stronger the dependency relationship between tasks. The strong dependency between the input task and the output task means that the input task is a predecessor task and the output task is a successor task.

For example, the number "9" in the 6th row and 4th column of DSM
Indicates a dependency relationship from the input task “design A” to the output task “design C”, which means that the degree of dependency is 0.9. That is, "design A" followed by "design C"
Is likely to continue.

As described above, in the DSM, complicated dependency relationships between tasks are simply expressed in a matrix format to facilitate risk management of product development projects and the like.

FIG. 3 shows DSM data as a result of rearranging the task order by the function of DSM called Partitioning.

In Partitioning, the task order is rearranged so that the numerical elements in the upper triangular matrix portion (the portion above the diagonal line) of the dependency relation expression unit 11 are reduced as much as possible. For example, the task order of the DSM shown in FIG. 2 is replaced with the task order shown in the DSM of FIG. That is, the DS of FIG.
In M, "outsourcing" is the input task unit 6 and the output task unit 7
In the DSM of FIG. 3, the input task unit 9 and the output task unit 10 are located at the 6th position.

As the task order is rearranged, the positions of the numbers included in the dependency relation expressing section 11 also change. Partitio
In ning, the task order is rearranged so as to reduce the number of numeric elements in the upper triangular matrix portion of the dependency relation expression unit 11.

The upper triangular matrix portion of the dependency relation expression portion 8 of FIG. 2 includes one "9", "5", and "1", respectively, while the dependency relation expression of FIG. In the upper triangular matrix portion of the part 11, "5", "3", and "1" are 1
Included individually. That is, the numeral element "9" in the upper triangular matrix portion of the DSM of FIG. 2 is moved to the lower triangular matrix portion of the DSM of FIG.

Assuming that the tasks are executed in the DSM data of FIG. 3 in the order from left to right of the input task section 9 or in the order from top to bottom of the output task section 10, the dependency relationship is obtained. It can be understood that the lower triangular matrix portion (the portion below the diagonal line) of the expression unit 11 represents the precedence relation of the task, and the upper triangular matrix portion represents the backward relation of the task. The shaded blocks of the dependency relationship expression unit 11 indicate the range of influence of the backtracking relationship.

According to the above assumption, when the tasks that can be executed are sequentially arranged and the dependency relationships between the tasks are connected by arrows, the flowchart shown in FIG. 4 is generated.

Box 12 represents a task, which is executed from top to bottom. A solid arrow 13 represents a task preceding relationship, and a dotted arrow 14 represents a task backward relationship. The numbers attached adjacent to the arrows 13 and 14 represent the degrees of dependence.

For example, the arrow from "outsourcing" to "manufacturing" is the precedence relationship of tasks, and the arrow from "design D" to "product planning" is the backward relationship of tasks. The arrows 13 and 14 representing the preceding relationship or the backward relationship of tasks are shown in FIG.
Corresponds to the number element of the dependency expression part 11 of the DSM.

[Task work period / execution order] FIG.
It represents the work period information of the task.

In each row of the table, work period information 16 of the task corresponding to the task 15 is shown. For example, the work period of the task “product planning” is 2 weeks. Here, the work period information of the task is not included in the DSM of FIG.

FIG. 6 shows an example of the task execution order determined based on the elements of the lower triangular matrix part (task precedence relation) of the DSM.

The length of the arrows 17 and 18 representing the task corresponds to the work period of the task. The working period of the whole project is 11 weeks. At this time, the tasks 18 indicated by white arrows, that is, “product planning”, “design B”, “outsourcing”, and “manufacturing” represent the connection of tasks with the least margin. This is called the critical path. In order to keep or shorten the work period of the entire project, it is necessary to focus the work on the critical path.

[Calculation of Task Execution Order and Critical Path] FIG. 7 shows the execution order (completion time) and critical path of each task based on the precedence relationship between tasks represented by DSM and the task work period. 3 shows an algorithm for calculating.

That is, the execution order (completion time) of each task and the critical path as shown in FIG. 6 are obtained by this algorithm.

Starting from the starting point 19, first, in a process 20, variables LIST_1, LIST_2, TIME (T) and C_PATH are initialized. That is, LIST_1 = {absent}, LIST_2 = {absent}, TIME (T) = 0, C_PATH = {absent}.

Further, as a pre-process for the calculation of the critical path, all the numerical elements of the upper triangular matrix part (return-back relation) of DSM are erased. This means that in the calculation of the critical path, the backward relationship is not considered, and only the task preceding relationship is considered.

Next, in the conditional branch 21, the variable LIST_
It is determined whether 1 matches all tasks of DSM. At the stage of moving from the process 20 to the process 21, the variable LIST_1
Does not include a task, the variable LIST_1 is
Since it does not match all the tasks of the DSM, the process 22 is proceeded to.

In the loop of processing 21 to processing 24, the execution order of each task, that is, the completion time of each task is determined.

In the first first loop, first, the process 2
In step 2, "task T having no input task" is obtained. For example, assuming that there are “product planning”, “design A”, “design B”, “design C”, “design D”, “outsourcing”, and “manufacturing” as tasks, “input task exists”. The task T ”that is not performed becomes“ product planning ”.

Then, this task is added to LIST_2 and L
IST_2 = {product plan}.

Next, in process 23, the completion time TIME (T) is set for the task T of LIST_2. TIME
(T) is the work period of the task T and the completion time T of the input task
It is set to the total value of IME (T) (or the maximum value of multiple completion times TIME (T) when there are multiple input tasks).

Here, since the task T which is an input task does not exist in the “product plan”, the completion time TIME (product plan) of the “product plan” is within the work period of the “product plan” (2 weeks). Will be equal.

Next, in the process 24, LIST_ is added to LIST_1.
Add task 2 and reinitialize LIST_2. That is, LIST_1 = {product plan} and LIST_2 = {none}. After that, the process returns to the conditional branch 21.

In the conditional branch 21 again, the variable LIST_
It is determined whether 1 matches all tasks of DSM. Since the variable LIST_1 includes only “product planning”, the variable LIST_1 does not match all DSM tasks, and the process proceeds to step 22.

In the second and subsequent loops, first, the processing 22
In ", a task T having no input task" or a "task T in which all input tasks are included in LIST_1" is obtained. For example, when LIST_1 = {product plan}, "task T in which all input tasks are included in LIST_1", that is, task T whose input task is "product plan" is "design A".
And "Design B".

Then, this task is added to LIST_2 and L
IST_2 = {design A, design B}.

Next, in process 23, the completion time TIME (T) is set for each task T of LIST_2. TIM
E (T) is the work period of task T and the completion time of the input task
It is set to the total value of TIME (T) (or the maximum value of a plurality of completion times TIME (T) when there are multiple input tasks).

Therefore, the completion time TIME (design A) of "design A" is the completion time (2) of "product planning" which is an input task.
Week) and the total value of the work period (3 weeks) of “Design A”,
That is, it will be 5 weeks. Also, the completion time TIME for "Design B"
(Design B) is the total value of the completion time (2 weeks) of the “product planning” that is an input task and the work period (4 weeks) of “Design B”, that is, 6 weeks.

Next, in process 24, LIST_ is added to LIST_.
Add task 2 and reinitialize LIST_2. That is, LIST_1 = {product plan, design A, design B}, and LIS
T_2 = {none}. After that, the process returns to the conditional branch 21.

In the conditional branch 21 again, the variable LIST_
It is determined whether 1 matches all tasks of DSM. "Product planning, design A, design B" included in variable LIST_1
Is not all DSM tasks, the process proceeds to process 22 (third loop).

First, in process 22, "task T with no input task" or "all input tasks are LIST_1".
The task T "included in is calculated. For example, LIST_1 =
When {Product planning, Design A, Design B}, "task T in which all input tasks are included in LIST_1", that is, task T whose input task is "product planning" or "design A" or "design B" Are “design C” and “outsourcing”.

Then, this task is added to LIST_2 and L
IST_2 = {design C, outsourcing}.

Next, in process 23, the completion time TIME (T) is set for each task T of LIST_2. TIM
E (T) is the work period of task T and the completion time of the input task
It is set to the total value of TIME (T) (or the maximum value of a plurality of completion times TIME (T) when there are multiple input tasks).

Therefore, the completion time TIME (design C) of "design C" is the total value of the completion time (5 weeks) of "design A" which is the input task and the work period (1 week) of "design C". In other words, the completion time TIME (outsourcing) of “outsourcing” is the completion time (6 weeks) of “design B” which is the input task and the work period (2 weeks) of “outsourcing”. It is the total value, that is, 8 weeks.

Next, in process 24, LIST_ is added to LIST_.
Add task 2 and reinitialize LIST_2. That is, LIST_1 = {product plan, design A, design B, design C,
Outsourcing} and LIST_2 = {none}. After that, the process returns to the conditional branch 21.

In this way, that is, in steps 21 to 24,
Repeat until variable LIST_1 matches all DSM tasks. That is, process 2 until LIST_1 = {product plan, design A, design B, design C, design D, subcontracting, manufacturing}.
Repeat 1-24.

In this example, the completion time TIME (design D) of "design D" is obtained in the fourth loop, and the completion time TIME (production) of "production" is obtained in the fifth loop. .
In the fifth loop, the “manufacturing” input task
There are two, "design D" and "outsourcing".
`` Outsourcing '' completion time TIME when seeking ME (manufacturing)
(Outsourced) is used.

When the variable LIST_1 matches all the tasks of DSM, it means that the execution order (or the completion time) of all the tasks has been decided, and thereafter, the processing advances to processing 25.

In processing 25-28, the task T that constitutes the critical path is obtained.

First, in process 25, the task T having the maximum completion time TIME (T) is obtained.

The task T having the maximum completion time TIME (T) means the task T to be executed last. For example, “manufacturing (TIME (manufacturing) = 11 weeks)” corresponds to this. .

Then, in process 26, this task T
To the list of critical paths, C_PATH,
= {Manufacturing}.

In processing 26 to processing 28, after the last task T to be executed that constitutes the critical path is obtained, the critical path is sequentially changed from the task T to the first task T to be executed. The constituent tasks T are specified.

In the conditional branch 27, a critical
It is determined whether or not the task T that is executed last in the path has a predecessor task.

When the last executed task T has no preceding task, that is, the completion time TIME (T) of the task T
, The critical path is composed of only the task T, and at the end point 29, the step of calculating the completion time and critical path of each task ends.

If the task T to be executed last has a predecessor task, that is, if the completion time TIME (T) of the task T does not match the work period, the process proceeds to step 28. In process 28, TIME (Ti) is selected from at least one task Ti that is an input task of the task T to be executed last.
= Task Ti that satisfies TIME (T)-(work period of task T)
And T = Ti.

Then, in process 26, this task T
To the list of critical paths C_PATH.

For example, as an input task of "manufacturing",
There are "design D" and "outsourcing", TIME (design D) = 7 weeks,
TIME (outsourcing) = 8 weeks. On the other hand, TIME (manufacturing) =
It is 11 weeks, and the working period of "manufacturing" is 3 weeks.

Therefore, the task Ti satisfying TIME (Ti) = TIME (manufacturing) − (working period of “manufacturing”) is “outsourcing”. Then, this "outsourcing" is added to the critical path list C_PATH, and C_PATH = {outsourcing, manufacturing}.

By repeating the loop of processing 26-28, for example, C_PATH = {product plan, design B, outsourcing, manufacturing} is obtained.

Here, the "product planning" to be executed first
Is finally added to the list of critical paths C_PATH. "Product planning" is a list of critical paths C_
When added to PATH, in process 27, T = “product planning”
Therefore, TIME (product planning) = "product planning" work period (2 weeks for each).

Therefore, the preceding task of the task T does not exist, and at the end point 29, the process of calculating the completion time and the critical path of each task ends.

As described above, according to the algorithm of FIG. 7, the completion time of each task and the critical path (ie,
It is possible to easily obtain a connection relationship between tasks as shown in FIG.

[Method of Quantifying Risk by Backtracking] FIG. 8
Is the result obtained by the algorithm of FIG.
A new DSM completed by adding to SM is shown.

The tasks composing the rows of the DSM 30 compose a plurality of bands 31. For example, one band is composed of one line of "product planning", one band is composed of two lines of "design A" and "design B", and "design C".
And two lines of "outsourcing" make up one band,
One row of "design D" constitutes one band, and one row of "manufacturing" constitutes one band.

Each band shows the contents of LIST_2, which is updated every time the loop is rotated, in the loop formed by the conditional branch 21, the process 22, the process 23 and the process 24 in the flowchart of FIG. That is, the first one
In the second loop, LIST_2 = {product planning} and 2
LIST_2 = {design A, design B} in the third loop, and LIST_2 = {design C, outsourcing} in the third loop.
Then, in the fourth loop, LIST_2 = {design D}, and in the fifth loop, LIST_2 = {manufacturing}.

In this case, the completion time TIME (T) of each task T is obtained by repeating the loop composed of the conditional branch 21, processing 22, processing 23 and processing 24 in the flowchart of FIG. 7 five times.

In the DSM of FIG. 3, no data was written in the diagonal line portion of the dependency expression part 11, but in the DSM of FIG. 8, the project start was written in the diagonal line part of the dependency expression part. The completion time TIME (T) of each task T when the time is set to “0” is written respectively.

Among the numbers in the diagonal line part of the dependency expression part of the DSM of FIG. 8, the white number 33 indicates the critical
It represents the completion time TIME (T) of the task T that constitutes the path C_PATH. For example, "outsourcing" is one of the tasks that make up the critical path, and the completion time TIME of "outsourcing" when the project start time is "0".
(Outsourcing) is 8 weeks.

Attention is now focused on the upper triangular matrix portion of the DSM 30, that is, the backtracking step.

In the calculation of the critical path, no information on the backtracking process of the task is referenced. But,
In fact, the DSM 30 contains three backtracking relationships. The first is the relationship of 0.1 (depending on the weight of DSM divided by 10) from "design D" to "product planning", and the second is from "outsourcing" to "design B". Dependency 0.
The third relationship, and the third is the relationship of the dependency degree 0.5 from “design C” to “design A”.

If the DSM 30 according to the present invention is used,
Risk management that considers the “return impact” can be easily performed.

The “return effect degree”, which is a quantification of the return risk, is defined as follows.

Predecessor task Ti is input task, successor task Tj
In the backtracking Ri, j whose output task is
Let d (i, j). Let t (i) be the completion time of task Ti, t (j) be the completion time of task Tj, and L (i) be the work period of task Ti.
Backtracking Ri, j has a backtracking impact on the project A (i, j)
Is A (i, j) = d (i, j) · {(t (j) −t (i)) + L (i)} (1).

It should be noted that (t (j) -t) on the right side of the above equation (1).
(i)) + L (i) represents the backtracking period.

For example, in the case of obtaining the backward influence degree A (i, j) from the subsequent task “design D” to the preceding task “product planning”, the “design D” and “product planning” are immediately obtained from the DSM 30 of FIG. Dependency d (i, j) = 0.1, "Product Planning" completion time t (i) = 2, "Design D" completion time t (j) =
7. The work period L (i) = 2 for "product planning" can be obtained.

Therefore, the degree of return from "design D" to "product planning" is A (i, j) = 0.1 {(7-2) +2} = 0.7.

Similarly, the backward influence degree from “outsourcing” to “design B” is 1.8 (dependency = 0.3, backward return period = 6), and from “design C” to “design A”. The regressive effect degree is calculated to be 2.0 (dependency = 0.5, relapse period = 4).

As described above, when risk management is performed on the basis of the index of the backward influence, it is understood that the backward influence degree from “design C” to “design A” has the greatest influence on the project. Therefore, it is possible to manage the risk of the project by taking measures such as allocating an excellent engineer in the processes of “design A” and “design C”.

By the way, the dependency between tasks, that is, DS
When determining the weight of M data, if the backtracking probability is used as a reference, from the above equation (1), the backtracking impact is obtained by multiplying the difference in completion time between tasks by the backtracking probability. Can be interpreted as representing the expected “delay” of the project.

That is, such a risk management can be evaluated more objectively than the conventional simple risk management using the FMEA RPN.

[Method of Quantifying Risk by Parallelization] For tasks having a dependency relationship, it is general to start the process of the subsequent task after the process of the preceding task is completed.

However, in order to shorten the construction period, there is a case where the process of the subsequent task is started by preparing for a certain risk before the process of the preceding task is completed. The risk in this case is that the work already performed by the succeeding task is wasted when it is found that the preceding task has a problem at the end of the process of the preceding task.

A method of quantifying the risk due to such parallelization will be described below.

The process chart of FIG. 9 is obtained by parallelizing some of the steps of the process chart of FIG. 6 while maintaining the preceding relationship of the process chart of FIG.

The parts 34 and 36 surrounded by the broken line are parallel processes.

In the parallel process 34, the construction period (working period) of "design A" is 3 weeks, and the construction period of "design C" is
It's a week. If the work of "design A" is started two weeks after the work of "design A" is started, the process of "design C" is also finished in the fifth week when the process of "design A" is finished. . In other words, the parallelization of tasks may shorten the construction period.

However, at the same time, the risk of parallelization must be taken into consideration. For example, when it is found that there is a problem in "design A" at the time when the process of "design A" is completed, the process of "design C" as well as the process of "design A" is redone. There must be.

In the parallel process 36, the design B construction period (working period) is 4 weeks, and the subcontracting construction period is 2 weeks. If the “outsourcing” work is started two weeks after the “design B” work is started, the “design B” process is completed 6
In the week, the "outsourcing" process is also completed at the same time. However,
For example, when it is found that there is a problem in "design B" at the time when the process of "design B" is completed, "design B"
In addition to the above process, the “outsourcing” process must be redone.

The construction period of the entire project in the process chart of FIG. 9 is the ninth week, which is shorter than the construction period of the entire project in the process chart of FIG. 6 by two weeks.
By parallelizing the tasks, the effect of shortening the construction period can be obtained, but on the other hand, the risk (delay of the construction period) when a problem occurs in the process of a certain task is large in proportion to the magnitude of the effect. Become.

In the process diagram of FIG. 9, the route consisting of “product planning”, “design B” and “manufacturing” is a critical path. Further, for example, if the work of "outsourcing" is started after a period of more than two weeks after the work of "design B" is started, "outsourcing" is included in the critical path.

As described above, when the tasks are parallelized, the project construction period is shortened, but the risk when a problem occurs in the task process is increased.

"Parallel influence degree" which quantifies the risk of parallelization is defined as follows.

Predecessor task Ti is output task, successor task Tj
In the parallelized Fi, j whose input task is, the dependency is d (i, j), the completion time of task Ti is t (i), the completion time of task Tj is t (j), task Tj Let L (j) be the work period of H, the parallel impact H (i,
j) is H (i, j) = d (i, j) · {t (i) − (t (j) −L (j))} (2).

It should be noted that t (i)-(t
(j) -L (j)) represents the overlapping parallel period of two tasks in the parallel process.

FIG. 10 is a DSM representation of the parallelization of the tasks introduced in FIG.

Compared with the DSM of FIG. 8, this DSM is
Due to the parallelization, a part of the task completion time (a part of the numbers in the diagonal line part of the dependency expression part of DSM) is changed.

The degree of parallel influence is calculated based on the lower triangular matrix portion of DSM.

For example, when the parallel influence degree H (i, j) of the preceding task "design B" and the succeeding task "outsourcing" is obtained, the dependency degree between "design B" and "outsourcing" is immediately obtained from DSM in FIG. d
(i, j) = 0.9, completion time of "design B" t (i) = 6,
"Outsourcing" completion time t (j) = 6, "Outsourcing" work period L
(j) = 2 can be obtained.

Therefore, the degree of parallel influence between “design B” and “outsourcing” is H (i, j) = 0.9 · {6- (6-2)} = 1.8.

Similarly, the parallel influence degree of "design A" and "design C" is calculated as 0.9 by multiplying the dependency degree (0.9) by the overlapping period (1w) of tasks. .

As described above, when risk management is performed based on the index of parallel influence, it can be seen that the parallel influence of “design B” and “outsourcing” has the greatest influence on the project. Therefore, it is possible to manage the risk of the project by arranging an excellent engineer in the process of “design B” and using an excellent “outsourcing”.

By the way, the dependency between tasks, that is, DS
When determining the weight of M data, if the output task failure probability (probability of occurrence of a problem) is used as a reference, from the above equation (2), the parallel influence degree is the output task failure probability in the task overlapping period. Therefore, it can be interpreted that the degree of parallel influence represents the expected value of “delay” of the project.

In other words, such risk management enables more objective evaluation than the conventional simple risk management using the FMEA RPN.

By the way, the back-back influence degree in the case of including the parallel process is also obtained by substituting the numerical value obtained from the DSM of FIG. 10 into the above equation (1).

For example, in the case of obtaining the backward influence degree A (i, j) from the subsequent task “outsourcing” to the preceding task “design B”, the degree of dependency from “outsourcing” to “design B” is calculated from DSM in FIG.
d (i, j) = 0.3, "Design B" completion time t (i) =
6. The completion time t (j) = 6 for “outsourcing” and the work period L (i) = 4 for “design B” can be obtained.

Therefore, the backward influence degree from “outsourcing” to “design B” is A (i, j) = 0.3 {(6-6) +4} = 1.2.

In the same manner, the backward influence degree from “design C” to “design A” is 1.5 (dependency = 0.5, backward return period = 3), and from “design D” to “product planning”. The regressive influence degree is calculated as 0.6 (dependency = 0.1, relapse period = 6).

In this way, when risk management is performed based on the index of the back-back influence degree in the case of including parallel processes,
It can be seen that the degree of backtracking impact from “design C” to “design A” has the greatest impact on the project.
Therefore, it is possible to manage the risk of the project by taking measures such as allocating an excellent engineer in the processes of “design A” and “design C”.

As described above, the risk management of the project including the parallel process can be quantified from the viewpoint of the backward influence degree and the parallel influence degree, and the objective risk evaluation can be performed.

[0143]

According to the present invention, the conventional FME based on the detailed process model represented by DSM data is used.
The risk of the backtracking process and the risk of the parallel process can be quantified more accurately than the risk analysis by A. As a result, for tasks on the critical path, it is possible to understand how delays due to backtracking processes and parallel processes affect the project, and
Compared with project management by path only, it is possible to manage time and risk more realistically.

[Brief description of drawings]

FIG. 1 is a diagram showing an apparatus for executing a project risk management method of the present invention.

FIG. 2 is a diagram showing a DSM in which dependency relationships between tasks are expressed in a matrix format.

FIG. 3 is a diagram showing a DSM in which task order is rearranged by partitioning.

FIG. 4 is a flow chart showing a preceding relationship and a backward relationship of tasks.

FIG. 5 is a diagram showing work period information of a task.

FIG. 6 is a diagram showing a critical path of a project.

FIG. 7 is a flowchart showing an algorithm for generating a task completion time and a critical path.

FIG. 8 is a diagram showing a DSM expressing a task completion time and a critical path.

[Fig. 9] Project criticality including parallel processes
Diagram showing paths.

FIG. 10 is a diagram showing a DSM of a project including parallel processes.

[Explanation of symbols]

1: Project management system 2: Display device 3: Database 4: Computing device 5: Interface 6,9: Input task unit 7,10: Output task unit 8, 11: Dependency display unit, 12 : Task, 13: Preceding relationship between tasks, 14: Backtracking relationship between tasks, 15: Task entry column, 16: Task work period entry column, 17: Task (not included in critical path), 18: Task ( (Included in critical path), 19: start point, 20, 22-26, 28: processing, 21, 27: conditional branch, 29: end point, 30: DSM, 31: band, 32, 38: completion of task Time (Critical
33, 39: Task completion time (critical, not included in path)
(Included in the path), 34, 36: Parallel process in which tasks with dependencies are parallelized, 35: Critical path of the project including the parallel process, 37: DS of the project including the parallel process
M data, 40: DSM weight.

Claims (17)

[Claims] 1. A first step of creating a design structure matrix in which a degree of dependency indicating a dependency between a plurality of tasks included in a project is expressed in a matrix format, and a work period and a preceding relationship of the plurality of tasks. On the basis of,
A second step of calculating a completion time of each task based on the start time of the project; and a start time of a second task preceding the first task from the completion time of the first task among the plurality of tasks And a fourth step of calculating a backtracking influence degree by multiplying the backtracking period by the degree of dependence from the first task to the second task. The characteristic project management method. 2. A first step of creating a design structure matrix in which a degree of dependency indicating a dependency between a plurality of tasks included in a project is expressed in a matrix format, and a work period and a preceding relationship of the plurality of tasks. On the basis of,
A second step of calculating a completion time of each task based on a start time of the project; a first task among the plurality of tasks; and a dependency on the first task, A third step of calculating a parallel period in which the first and second tasks overlap with respect to a second task having a parallel relationship that starts before the task is completed; and a degree of dependence from the first task to the second task And a parallel step, the fourth step of calculating a parallel influence degree by multiplying the parallel period. 3. The project management method according to claim 2, wherein a fifth step of calculating a backtrack period from the completion time of the second task to the start time of the first task, and the second task to the first A sixth step of calculating the degree of backtracking influence of the first and second tasks having the parallel relationship by multiplying the degree of dependency on the task and the backtracking period. 4. The project management method according to claim 1, wherein the completion times of the plurality of tasks are recorded in a diagonal line portion of the design structure matrix. 5. The completion time of the tasks constituting the critical path among the completion times of the plurality of tasks is recorded separately from the completion times of other tasks. Described project management method. 6. The plurality of tasks are arranged in the design structure matrix in ascending order of start time, and a triangular matrix portion on one side of the diagonal line portion has a preceding relationship, and a triangular matrix portion on the other side. 5. The project management method according to claim 4, wherein represents a backtracking relationship. 7. A program for executing the first to fourth steps according to claim 1 or 2. 8. A computer-readable recording medium in which a program for executing the first to fourth steps according to claim 1 or 2 is recorded. 9. A design structure matrix in which the degree of dependency indicating a dependency among a plurality of tasks included in a project is expressed in a matrix format, a database in which a work period and a preceding relationship of the plurality of tasks are recorded, and An arithmetic unit for calculating the completion time of each task based on the start time of the project based on the work period and the preceding relationship, wherein the arithmetic unit is the first task among the plurality of tasks. From the completion time to the start time of the second task preceding the first task, and multiplying the degree of dependence from the first task to the second task by the backtrack period. A project management device characterized by calculating a backtracking influence degree. 10. A design structure matrix in which a degree of dependency indicating a dependency between a plurality of tasks included in a project is expressed in a matrix format, a database in which a work period of the plurality of tasks and a preceding relationship are recorded.
An arithmetic unit for calculating the completion time of each task based on the start time of the project based on the work period and the preceding relationship, wherein the arithmetic unit is the first of the plurality of tasks. For a task and a second task having a dependency on the first task and having a parallel relationship that starts before the first task is completed, a parallel period in which the first and second tasks overlap is calculated. Further, the project management apparatus is characterized in that the degree of parallel influence is calculated by multiplying the degree of dependence from the first task to the second task and the parallel period. 11. The project management method according to claim 10, wherein the arithmetic device calculates a backtracking period from the completion time of the second task to the start time of the first task, and calculates from the second task. A project management apparatus, wherein the degree of backtracking influence of the first and second tasks having the parallel relationship is calculated by multiplying the degree of dependence on the first task and the backtracking period. 12. The completion time of the plurality of tasks is recorded in a diagonal line portion of the design structure matrix.
Described project management device. 13. Among the completion times of the plurality of tasks,
13. The project management apparatus according to claim 12, wherein the completion times of the tasks forming the critical path are recorded separately from the completion times of other tasks. 14. The plurality of tasks are arranged in the design structure matrix in ascending order of start time, and a triangular matrix portion on one side with respect to the diagonal line portion has a preceding relationship, and a triangular matrix portion on the other side. 13. The project management apparatus according to claim 12, wherein represents a backtracking relationship. 15. A design structure matrix in which a dependency indicating a dependency between a plurality of tasks included in a project is expressed in a matrix format, wherein the plurality of tasks are input tasks in a row direction in ascending order of start time. As the output tasks in the column direction, they are arranged in the design structure matrix, respectively, and the completion time of each task based on the start time of the project is arranged in the diagonal line portion of the design structure matrix. The design structure matrix is characterized in that, with respect to the diagonal line portion, a triangular matrix portion on the input task side represents a backward relationship and a triangular matrix portion on the output task side represents a preceding relationship. 16. Among the completion times of the plurality of tasks,
16. The design structure matrix according to claim 15, wherein the completion times of the tasks constituting the critical path are recorded separately from the completion times of other tasks. 17. The plurality of tasks as the output tasks are divided into bands according to a preceding relationship of the plurality of tasks, and a work period of the plurality of tasks is a preceding task that precedes a completion time of a succeeding task. 2. The period is equal to the period obtained by subtracting the completion time of the task.
The design structure matrix described in 5.