JPH04348856A - Work starting order deciding system - Google Patents

Abstract

PURPOSE:To calculate (decide) the degree of starting priority (starting order) of each lot by allowing latitude in the completion scheduled data of each by using a process lead time in which latitude is allowed. CONSTITUTION:From a lead time in which latitude, such as shortmost, average, and longmost, is allowed and a present data, a completion date of each of lots A, B... classified by kind of part groups (a), (b)... is predicted with latitude, such as rapidmost, average, latemost, allowed. From a relation between the completion data and a delivery date, the degree of priority (the degree of emergency) is calculated, and a starting order is decided according to the degree of priority. The degree of priority is determined by (average - delivery time)/(latemost - average) when a delivery date of, for example, A is between average and latemost. The degree of priority is determined by (average - a delivery date)/(latemost - average) when delivery dates of B and C are between average and rapidmost. When a delivery data is before rapidmost, it is set to 1.0 and when a delivery data is after latemost, it is set to -1.0.

Description

[Detailed description of the invention]

0001

[Industrial Application Field] The present invention is directed to the production of a plurality of ready-to-start parts (lots) that can be completed by working in a shop (workplace, workshop) consisting of a plurality of mechanical devices (processes).
The present invention relates to a method for determining the starting order of these tasks, and particularly relates to a method for determining the starting order of these tasks so as to meet a delivery deadline.

0002

2. Description of the Related Art Conventionally, as a method for determining the starting order of work, a dispatching rule or a rule based on the know-how of experts has been used, for example, as described in Japanese Patent Laid-Open No. 2-212053. A method using an expert system has been devised. Here, the dispatching rule is applied to predict the completion date from the process lead time of each lot, and the possibility of meeting the delivery date is determined from the relationship between the predicted completion date and the delivery date.

0003

[Problems to be Solved by the Invention] In the above-mentioned prior art, the completion date is predicted from the process lead time of each lot fixed in advance by applying a dispatching rule, and the predicted completion date and delivery date are If they match or the predicted completion date is earlier than the delivery date, it is determined that the delivery date can be met; otherwise, it is determined that the delivery date cannot be met. However, because the lead time for each lot changes depending on the current workload of the relevant shop, the work in progress in the shop, the operating status of machinery, etc., the predicted completion time obtained from the fixed lead time It is inappropriate to judge the possibility of meeting deadlines by days (given as 1 point). For example, if the delivery date is 10 days from now and the completion date is predicted to be 11 days from now, it is impossible to determine whether or not the delivery date can be met by using express instructions. On the other hand, if the requested delivery date is 10 days later, but the completion date is predicted to be 8 days later, it is difficult to decide whether or not to start this lot now, and whether the delivery date can be met even when the shop's workload is heavy. It is also difficult to judge whether Therefore, there has been a problem in that delivery dates may be delayed depending on the workload of each shop, the amount of work in progress within the shop, the operating status of mechanical equipment, etc. Further, the method of creating a schedule using various dispatching rules and adjusting the schedule while evaluating the schedule requires a lot of calculation time.

[0004] Therefore, an object of the present invention is to provide a width to the process lead time in order to solve the above-mentioned problems of the prior art and to increase the possibility of meeting the delivery date. The object of the present invention is to provide a work start order determining method that makes predictions with a wide range and determines the work start order according to the relationship between the predicted completion date and the delivery date.

[0005]

[Means for Solving the Problems] In order to achieve the above object, the present invention provides a work start sequence for a plurality of lots (startable parts) that are completed by successively working in a plurality of processes (mechanical devices). In the work start order determination method that determines the work start order, there is a lot information storage unit that stores the type of parts and delivery date of each lot, and the shortest period from start to completion (process lead time) in the corresponding shop for each part type group. Using the process lead time storage section that stores the average and longest, and the shortest, average, and longest process lead times obtained from the process lead time storage section regarding the type of part and the current date, the completion date is stored as the earliest, average, It is equipped with a section completion date prediction unit that predicts based on the latest section (section completion date), and a start order determining section that determines the work start order of each lot based on the predicted section completion date (predicted completion date) and delivery date. It is.

[0006] Also, there is a process lead time record storage unit that stores past process lead time records for each part type, and a process lead time record storage unit that stores the shortest, average, and longest processes from the past records stored in the process lead time record storage unit. The system also includes a process lead time prediction section that calculates lead time.

[0007]Furthermore, in order to predict the process lead time in consideration of the workload, one or both of a work status storage unit that stores the current work status and a production plan storage unit that stores the production plan; It is equipped with a workload calculation unit that calculates the workload of the corresponding shop using information stored in the work status storage unit and the production plan storage unit.

[0008]

[Operation] The operation based on the above configuration will be explained.

According to the present invention, for each lot that can be started (for example, in the relevant shop), the value of the part type of each lot retrieved from the lot information storage section is used to store the process lead time storage section. The shortest time from start to completion of the stored part type (for example, in the relevant shop),
On average, the longest process lead time is taken. next,
Based on the shortest, average, and longest process lead times obtained earlier and the current date, the section completion date prediction unit predicts the completion date in the earliest, average, and latest sections (section completion dates). The start order determining section determines the work start order for each lot based on the predicted section completion date and the delivery date of each lot retrieved from the lot information storage section. For example, if the delivery date is before the earliest predicted completion date, the delivery date will be delayed, so the priority for starting work is set high.
If the delivery date is between the earliest predicted completion date and the average predicted completion date, it is possible to meet the delivery date, but since detailed management is required, the priority should be set slightly higher and the delivery date is between the average predicted completion date. If the delivery date is between the date and the latest predicted completion date, there is some margin for the delivery date, so set the priority slightly lower. If the delivery date is after the latest predicted completion date, there is no need to start yet, so set the priority. Set the priority based on your judgment, such as setting it low.
Based on this, the order in which work begins on each lot is determined.

[0010] The above method calculates the shortest, average, and longest process lead.
Fixed data stored in the process lead time storage unit is used as the process lead time, but process lead time can be calculated using past process lead time records stored in the process lead time performance storage unit. - It is also possible to use a lead time prediction unit to predict the shortest, average, and longest process lead times for the corresponding part type. This shortest, average, and longest process lead
A more accurate prediction can be made by predicting the completion date using the detime.

[0011]Furthermore, the workload calculation unit calculates the workload of the corresponding shop from information in either or both of the work status storage unit that stores the current work status and the production plan storage unit that stores the production plan. It is possible to predict the process lead time in consideration of the information and use this to more accurately predict the completion date.

0012

Embodiments Below, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 shows a schematic functional configuration of a work start order determining system according to an embodiment of the present invention. In FIG. 1, 101 is a lot information storage unit that stores the part type and delivery date of each lot, and 102 is a part that stores the shortest period from start to completion (process lead time) for each part type and shop.
Process lead time storage unit that stores average and longest time, 103
uses the shortest, average, and longest process lead times obtained from the process lead time storage and the current date to determine the earliest completion date.
A section completion date prediction section 104 predicts based on the average and latest section (section completion date), and a start order determining section 104 determines the work start order for each lot based on the predicted section completion date and delivery date. In addition, there is a process lead time record storage unit 111 that stores process lead time records by part type and shop, and the shortest, average, and longest process lead times from the work records stored in the process lead time record memory unit 111. - By using the process lead time prediction unit 112 for calculating lead times, the shortest, average, and longest process lead times can be determined from past work results. Furthermore, when predicting the process lead time by taking into account the workload, the information stored in the work status storage unit 121 that stores the current work status of each shop and the production plan storage unit 122 that stores the production plan. A workload calculation unit 123 is required to calculate the workload of the shop using the information.

FIG. 2 shows a work start order determination method according to an embodiment of the present invention in three shops (shop 1, shop 2,
This figure shows the overall system configuration when applied to a machining workplace consisting of shop 3). Each shop has work record collection terminals 201 to 20 such as barcode readers for collecting work records such as starting and completing lots.
3, and work instruction terminals 221 to 223 such as personal computers for displaying work start instructions to workers 211 to 213 are installed. These terminals are connected via a network 231 to a shop management computer 232 which centrally manages all shops. All the functions 101 to 123 of the work start order determining method shown in FIG. 1 are executed by the shop management computer 232.

The work start order is determined by the worker 2 of each shop.
This is executed when a request is made to display a work instruction from the work instruction terminals 221 to 223 when the robots 11 to 213 are about to start a new lot machining operation. Hereinafter, the work start order determination method will be explained assuming that there are three lots that can be started: lot A (parts type a), lot B (parts type b), and lot C (parts type c).

The section completion date prediction unit 103 predicts the completion date for each of the available lots according to the following procedure.

1) The type of parts and delivery date related to the lot are retrieved from the lot information storage section 101.

2) The shortest, average, and longest process lead times are retrieved from the process lead time storage unit 102 for each shop that performs the machining work of the corresponding lot.

3) Add together the shortest process lead times for each shop obtained in 2) above to determine the shortest process lead time until the corresponding lot is completed. Similarly,
Find the average and longest process lead time until the corresponding lot is completed.

4) Find the section completion date from the shortest, average, and longest process lead times found in 3) above and the current date.

For example, assume that lot A (part type a) is processed in the order of shop 1, shop 2, and shop 3, and the shortest, average, and longest process lead times for each shop in lot A are as follows:
Shortest Average Longest
Shop 1 10 days
12th 15th Shop 2 8th 9th 10th
Shop 3 1
If it is the 5th, 20th, and 26th, then shop 1
The shortest, average, and longest process lead times from the processing operations in Shops 2 and 3 to the completion of Lot A are the shortest process lead times.
33rd
Average process lead time 41
Days Longest process lead time
It will be 51 days. Here, the shortest process lead time means that if this lot is prioritized and processed, the waiting time in front of each machine will be shortened, so the process lead time can be shortened to this period. However, the longest process lead time means that if the priority of this lot is lowered and the machining work is performed, the waiting time in front of each machine will be longer, but the work will be completed within this period at the latest. Also, the average process lead time means that if the work progresses normally, it will be completed within this period on average.

Next, the shortest, average, and
The section completion date for each lot is predicted from the longest process lead time and the current date. The section completion date is actually predicted using the factory's actual working day calendar, such as the earliest predicted completion date, X month, X day, the average predicted completion date, Y month Y day, and the latest predicted completion date, Z month Z day. , In the explanation of this example, the earliest predicted completion date,
The average predicted completion date and latest predicted completion date are expressed as the cumulative number of days from the current date. Even if this is done, the concept of the present invention will not be affected in any way.

The start order determining unit 104 first compares the predicted section completion date for each lot with the delivery date of the corresponding lot retrieved from the lot information storage unit 101 to determine the start priority of each lot. In this embodiment, start priority is considered as follows.

(1) If the delivery date is before the earliest predicted completion date, the priority for starting work is set high because the delivery date will be delayed.

(2) If the delivery date is between the earliest predicted completion date and the average predicted completion date, the delivery date can be met, but detailed management is required, so the priority is set a little higher.

(3) If the delivery date is between the average predicted completion date and the latest predicted completion date, the priority is set slightly lower because there is some margin for the delivery date.

(4) If the delivery date is after the latest predicted completion date, the priority is set low because there is no need to start yet.

Various methods can be considered for assigning start priority, but in this embodiment, the priority is set in the range of 1.0 to -1.0 as shown below, and the start priority is set as follows. Lots with higher urgency are given higher priority.

In the case of (1), priority = 1.0 In the case of (2), the priority = (average predicted completion date - delivery date) / (average predicted completion date - earliest predicted completion date) In the case of (3), priority = (Average predicted completion date - delivery date) / (latest predicted completion date - average predicted completion date) In the case of (4), priority = -1.0 In this way, the priority of each lot is determined and ordered in descending order of priority. By determining the order in which the work is to be started, instructions for starting the work for each lot can be given in consideration of compliance with the delivery deadline.

[0030] A method for determining priorities will be explained using an example. Assuming that the delivery date of lot A is 45 days from the contents of the lot information storage unit 101, the earliest predicted completion date of lot A is 33 days, the average predicted completion date is 41 days, and the latest predicted completion date is 5 days.
Since it is one day, the delivery date of lot A corresponds to case (3), which is between the average predicted completion date and the latest predicted completion date, and the priority is -0.4. Similarly, find the shortest, average, and longest process lead times until completion for lots B and C, and
Lot B Lot C
Earliest expected completion date: 18th
15 days Average expected completion date
20th 20th
Latest expected completion date: 22nd
25 days is obtained, and the delivery dates for both lots B and C are 19 days.In case (2), when both delivery dates are between the earliest predicted completion date and the average predicted completion date, Correspondingly, the priorities of lot B and lot C are 0.5 and 0.2, respectively. Based on the above results, a work start instruction is given to start work on lot B, lot C, and lot A in this order. In the normal method, it is difficult to determine which lot should be started first because the average expected completion date and delivery date for Lot B and Lot C are the same, but according to the example, By determining the predicted date by interval, determining the priority based on this, and determining the work start order so that the work starts in descending order of priority, it is possible to give work start instructions that take delivery deadlines into consideration. .

FIG. 3 shows a work start instruction screen displayed on the work instruction terminals 221 to 223. Along with the lot name, part type, and delivery date of available lots, the earliest predicted completion date, average predicted completion date, latest predicted completion date, and priority are displayed. The operators 211 to 213 can easily decide which lot to start work on.

In the above explanation, a case has been described in which fixed data stored in the process lead time storage unit 102 is used as the shortest, average, and longest process lead times. In order to effectively use the order determination method, it is necessary to appropriately set the values of the shortest, average, and longest process lead times without fixing them. Therefore, a method for determining the shortest, average, and longest process lead times using past process lead time records will be described.

[0033] In order to find the shortest, average, and longest process lead times from the actual results, the work results for the start and completion of each lot for each shop are collected on work result collection terminals 201 to 203 such as barcode readers. The information is collected by the process lead time record storage unit 111, transferred to the shop management computer 232 via the network 231, and stored in the process lead time record storage unit 111. Process lead for each lot
The time performance can be easily determined by subtracting the start date of the lot from the completion date of the lot. Process Lee
The lead time prediction unit 112 calculates the shortest and longest process lead times for each part type and shop from the process lead time records retrieved from the process lead time record storage unit 111 for each fixed period such as days or weeks. , respectively, are set as the shortest and longest process lead times, and the average process lead time is calculated using ordinary arithmetic mean, and this result is stored in the process lead time storage unit 102. , the latest shortest, average, and longest process lead times can be set.

The process lead time is also affected by the amount of work load in the shop, the work in progress in the shop (work that has not been completed after it has started), the operating status of machinery, and the like. Therefore, a method for determining the shortest, average, and longest process lead times will be explained by taking into account the workload of the shop, the amount of work in progress in the shop, the operating status of machinery, and the like. In the following explanation, we will discuss a method that takes into account only the in-process work in the shop, but if we use the same concept, we can calculate the shortest, average, and longest processes for each lot, taking into account the workload, operating status of machinery, etc. Lead time can be predicted.

[0035] Process lead time is affected by the work in progress within the shop, and as the work in progress within the shop increases, the process lead time decreases.
The lead time becomes longer, and conversely, as the amount of in-process work in the shop decreases, the process lead time becomes shorter. Therefore, the relationship between the work in progress within the shop and the process lead time is determined using the work results in the process lead time performance storage unit 111, and when predicting the process lead time, the relationship between the work in progress within the shop and the process lead time is calculated using Let us consider predicting the process lead time using the relationship between the work in progress in the shop and the process lead time, which has been determined in advance. Hereinafter, an example will be explained in which the relationship between the in-process work in the shop and the process lead time for part type a and shop 1 is determined.

For all lots related to part type a and shop 1 stored in the process lead time record storage unit 111, the relationship between the work in progress in the shop and the process lead time is determined. The case of lot A, which satisfies the above conditions, will be explained below as an example. First, find the number of work in progress when lot A starts. A lot whose start date is before the start date of lot A and whose completion date is after the start date of lot A is a lot that is in progress in the shop when lot A starts.

FIG. 4 shows a part of the structure and contents of the process lead time record storage unit 111 regarding shop 1. The start date of lot A is 2/1, and the start date of lot B is 2/1. Since Lot B is 1/25 and the completion date is 2/8, Shop 1 is in progress at the time Lot A starts. Similarly, it can be seen that Lots D and F are in progress in Shop 1 when Lot A starts. That is, the number of work in progress at the time of starting Lot A is 3. Similarly, the number of works in progress for shop 1 at the time of starting lot A is calculated for all work records for shop 1 (regardless of part type) stored in process lead time record storage unit 111. Now lot A
The number of work in process in shop 1 at the time of start is 30 and lot A
If the lead time of is 18, then part type a,
Regarding Shop 1, we have obtained a track record that the process lead time is 18 days when the number of work in progress is 30. Similarly, all lots with part type a and already completed (with a completion date) (lot D in Figure 4)
and E apply. In reality, many lotteries are stored as past results. ), the relationship between the in-shop work in progress and the process lead time for shop 1 is calculated, and this relationship is expressed in a graph as shown in FIG. Here, individual lots are marked with an x. From this graph, there are several ways to find the relationship between in-process work in the shop and the shortest, average, and longest process lead times for part type a and shop 1. If used, as shown in Figure 6, the relationship between the number of work in progress in the shop (W) and the shortest, average, and longest process lead times (Lmin, Lmean, Lmax) is determined by an interval regression equation, Lmin=Kmin× W+Dmin Lmean=Kmean×W+DmeanLmax=K
Coefficients Kmin, Dmin, Kme to calculate as max×W+Dmax
an, Dmean, Kmax, and Dmax can be calculated.

Since the current number of work-in-progress (W) in the shop regarding shop 1 is stored in the work status storage unit 121, the process lead time prediction unit 112 uses the above interval regression formula to estimate the number of work in progress for part type a. Shortest in shop 1,
The average and longest process lead time can be determined. For example, if the current number of work in progress (W) in shop 1 is 35, as shown in FIG.
The shortest, average, and longest process lead times are 13 each.
Day, 16th, 19th, etc.

As described above, the shortest, average, and longest process lead times for each lot can be predicted using the past process lead time records stored in the process lead time record storage unit. Therefore, it is possible to more accurately predict the completion date by reflecting actual results and taking into consideration the work in progress in the shop, the amount of work load, the operating status of machinery and equipment, etc.

[0040] In the above embodiment, there is a worker in each shop, and the work is carried out in accordance with the start instructions from the work instruction terminal, but if the shop is highly automated, The same effect can be obtained by directly instructing machines and peripheral equipment to start work instead of instructing them to start work from a work instruction terminal according to the start order determined by the start order determination method. is clear.

[0041]

Effects of the Invention As explained in detail above, according to the present invention, the completion date of each lot can be determined by using the process lead time and the current date, which have a range of widths such as earliest, average, and latest.
Forecast with a range of average and slowest, determine priority (urgency) from the relationship between these predicted completion dates and delivery dates, and determine the work wear procedure for each lot according to this priority. This has the effect of making it possible to make detailed, flexible and appropriate decisions regarding compliance with delivery dates, thereby increasing the possibility of compliance with delivery dates.

[Brief explanation of the drawing]

FIG. 1 is a functional configuration diagram showing schematic functions of an embodiment of the present invention.

FIG. 2 is a system configuration diagram of an entire system to which an embodiment of the present invention is applied.

FIG. 3 is a diagram showing a work start instruction screen according to an embodiment of the present invention.

FIG. 4 is a diagram showing the structure and contents of a process lead time record storage section.

FIG. 5 is a diagram showing the relationship between the number of in-process items in a shop and process lead time.

FIG. 6 is a diagram showing the relationship between the number of in-process items in a shop and the shortest, average, and longest process lead times.

[Explanation of symbols]

101 Lot information storage unit 102 Process lead time storage unit 103 Section completion date prediction unit 104 Start order determination unit 111 Process lead time performance storage unit 112 Process lead time prediction unit 121 Work status storage unit 122 Production plan storage unit 123 Work Load calculation section

Claims (3)

[Claims] Claim 1: In a work start order determination method that determines the work start order for multiple lots that are completed by successively performing work in multiple processes, a lot information storage that stores the type and delivery date of parts of each lot. Minimize the process lead time, which is the period from start to completion, by department and part type.
Using the process lead time storage unit that stores the average and longest process lead times, and the shortest, average, and longest process lead times obtained from the process lead time storage unit regarding the type of part and the current date, the completion date is determined as soon as possible. Work start is characterized by having a section completion date prediction unit that predicts based on average and latest section completion dates, and a start order determining unit that determines the work start order for each lot based on the predicted section completion date and delivery date. Ordering method. [Claim 2] A process lead time record storage unit that stores past process lead time records for each part type, and information in the process lead time record storage unit to determine the shortest, average, and longest process lead times. 2. The work start order determining method according to claim 1, further comprising a process lead time prediction unit that calculates the process lead time. 3. A work status storage unit that stores the current work status, a production plan storage unit that stores the production plan, and a corresponding process based on information from either or both of the work status storage unit and the production plan storage unit. 3. The work starting order determining method according to claim 2, further comprising: a workload calculation section for calculating the workload of the shop, and is configured to predict the process lead time by taking into account the workload information.