Pom Lecture (42)

Unit 3 Scheduling Operations Chapter 13: Scheduling Lesson 41 - OPERATING SCHEDULES Learning Objectives After reading this lesson you will be able to understand Operating schedule Conversion systems Loading Sequencing

Operating schedules When you come to the campus for studies, you have a daily workout of what classes on which subject are going to be held in what time frame. Well this is Scheduling of your courses. Similarly applying this analogy in industry whether you are making a product or giving service, the day to day planning of operations is called scheduling. Therefore in this chapter we will cover Operations schedule and the intermittent system. But first thing is first. Operating schedules are short-term plans even day-to day plans designed to implement the master schedule plans. We have seen earlier that once a business plan is in place then a master schedule is made showing how many of each product must be produced according to the customer orders and demand forecast. As discussed earlier, conversion system can be broadly classified as either continuous or intermittent depending on the conversion process and the product or service. A continuous or assembly type system is one in which a large number or infinite number of units of a homogenous product is produced.

An intermittent system, on the hand, produces a variety of products one at a time ( in which case they are custom made ) or in batches to customer order. Many conversion facilities are neither strictly intermittent nor continuous but a combination of both. Let’s see their implication in a manufacturing context. MANUFACTURING: In a manufacturing context, intermittent systems are traditionally referred as Job work or shops. Therefore a number of questions arise such as •

Which work centers will do which job?

•

When should an operation / job be started? When should it end?

•

On which shipment should be done, and by whom?

In the next part of the chapter we will attempt to answer the above questions. In manufacturing the sequencing in which waiting jobs are processed is critical to the efficiency and effectiveness of the intermittent systems. Sequencing affects how many jobs are completed on time versus late, costs incurred for set up and changeover, delivery lead times, inventory costs, and the degree of congestion in the facility scheduling of the intermittent systems poses a challenge for operations manager. At this point I would like to bring to your attention the fact that the intermittent systems also find extensive applicability in Services. Allow me to properly explain. SERVICES The intermittent systems can also be used in the service sector. The concept is similar in nature as manufacturing. INTERMITTENT SCHEDULING CONCEPTS AND PROCESSES In this section we will discuss about the various concepts of Scheduling and control process. The whole process can be shown in the Figure 13.1 below:

BUSINESS PLAN OPERATIONS Output Planning

Capacity Planning

Aggregate output planning

Aggregate Capacity Planning

Master Production Scheduling

Rough out Capacity Planning

Material Requirement Planning

Detailed Capacity Planning

Loading

Sequencing

Shop Floor Level

Detailed Planning

Expediting Figure 13.1 A logical question to ask would be:1. What is loading ?

Short Term Capacity ( input / output control )

Loading In a job shop operations we have many work centers. We also have customers orders. Each order has a unique specification and hence it has to be routed through various work centers in the facility. The new orders are assigned or allocated among the work centers. Then each work center has a certain load of various orders in the planning period. This assignment is known as Loading. Sometimes in, with the best scheduling there are imbalances, it has to be manipulated or managed to get the best output results. Two approaches for managing the imbalances are :Infinite loading and finite loading. Let me explain.

Infinite Loading Here the jobs are assigned to the work centers without regard to the work centers capacity; jobs are loaded from the production schedule into the work centers as if its capacity is infinite. Again there are three ways are available for evaluating the current loadings 1.

The Gnatt Load Chart.

This chart is used for sequencing and monitoring its

progress. The progress chart graphically displays the current status of each job relative to its completion date. For example, suppose that an automobile parts manufacturer as three jobs underway, one each for Ford, Plymouth and Pontiac. The actual status is shown by box.. As on date April 21, the Gnatt Chart shows that the Ford order lags behind schedule because operations has completed only the work scheduled through April 18. The Plymouth order is exactly on schedule and the Pontiac is ahead of schedule.

JOB

17/4

18/4

19/4

20/4

21/4

22/4

23/4

24/4

25/4

26/4

FORD PLYMOUTH PONTIAC

Start Activity ,

Finish Activity.

Scheduled Activity Time,

Actual Progress.

Another figure below shows a machine chart for the automobile parts manufacturer. This chart shows the sequence of future work at two machines and also can be used to monitor progress. Using the same notion as in the figure above, the chart shows that for the current of April 21, the Plymouth job is on schedule at the grinder because the actual progress coincides with the current date. The Pontiac order has finished at the lathe, which is now idle. The manager can easily see from the Gnatt machine Chart, the consequence of juggling the schedules. The usual approach is to juggle the schedules by trial and error until a satisfactory level of selected performance measurement is obtained. 2. Visual Load Profile

Here the profile compares the load and capacity of the work

center to avoid the work center being overloaded or under loaded

In a manual scheduling system , in figure on left hand side the load consists of

orders (

existing orders from customers) assigned to the work center. The load for the week 1 exceeds capacity but the future loads are well within the capacity. In the computer based scheduling system the right hand side of the picture the load consists of open orders and planned orders (prospective orders from customer). We see that the loads in the week 3,4 & 6 exceed capacity, even though loads for the open order are feasible. Dear students, now I’m introducing an extremely important concept, i.e. The Assignment Algorithm This is used for solving the loading problems. It is a linear program to assign jobs so that a specific criterion is optimized. Managers often have choices about which jobs should be assigned to which job centers or perhaps the best center is not available because it is already assigned to a job and can process one job at a time The method can be used when the number of jobs equals to the number of work centers or the machines the job requires. The method also requires that each machine be assigned one and one job at a time. Furthermore, the goodness or badness of the assignments must be specifiable by some quantifiable basis. The basis may be profit, operating costs or completion time. Thus, the method cannot be used if jobs are arriving continuously. Let’s focus on:-

FINITE LOADING It is a scheduling technique that combines into a single system the loading, sequencing and detailed scheduling. Here the system starts with a specified capacity for each work center and a list of jobs. The work center’s capacity is then allocated unit by unit (for e.g. Labor hours ) to the jobs by simulating job stating times and completion times. Thus the system creates a detailed schedule for each job and each work center based on the centers capacities. Jobs are allotted to the centers according to their capacities hour-by-hour and

day by ay into the future. The resultant finite capacity load would look like in the figure given below: Xxxxx

xxxxxxx xxxxxx xxxxxxx xxxxxx xxxxxx

Xxxxxx xxxxxxx xxxxxx xxxxxxx xxxxxx xxxxxx Xxxxxx xxxxxxx xxxxxx xxxxxxx xxxxxx xxxxxx xxxxxxx xxxxxx xxxxxxx

xxxxxx

xxxxxxx xxxxxx xxxxxxx

xxxxxx

xxxxxxx

xxxxxxx

xxxxxx

xxxxxxx

xxxxxxx

xxxxxx xxxxxx

The above figure shows Finite Capacity load Profile for work center when capacity is 100 labor hours. Open shop orders,

xxxx

Planned order.

So you see the working of the finite loading systems as compared to the infinite loading systems. Again you will notice that the sequencing rule into the simulation of infinite loading system. Since inputs to the simulator ( for e.g. from an MRP system ) specify due dates ( not starting dates or completion dates ) the jobs can be loaded using either forward or backward loading scheduling, as we see next. Friends, two pertinent types of scheduling should be discussed here. These are:Forward scheduling This system also known as set forward, is normally used in the job shops where customers place their orders on a “ needed as soon as possible “basis. Forward scheduling determines the start and finish times for the next priority jobs by assigning it to the earliest available time slot, and from that time, determines when the job will be finished in that work center. Since the job and its components start as early as possible, they will typically completed before they are due at subsequent work centers in the routing. As a

result this system accumulates in- process inventories that sit through out the facility until they are needed at subsequent work centers. On the other hand it is simple to use and it gets job done in shorter time. Backward Scheduling This system also known as set backward is often used in assembly type industries and in job shops that commit, in advance, to specific delivery dates. Backward scheduling assigns the next priority job the latest available time the latest available time slot that will enable the job to be completed just when it is due, but no before to the next work enter in the routing. Then the starting time is determined by “ setting back” from the finish date. By assigning jobs as late as possible, backward scheduling minimizes inventories since the job is not completed until it must go directly in the next work center in the routing. Thus in this system the inventory build up is avoided. However strict control of Bill Of Materials and the lead time estimates have to be done otherwise it may result in system breakdown, due dates violation and delivery service to the customers deteriorates. I guess we had enough of theoretical inputs. Let us try to understand the above concepts through a example given below: An Example A ZXS company has received two orders, A & B, both of which require processing at M/c 1 & 2. The first come first serve rule is applied to sequence the jobs, here job A arrived in advance of job B. The sequence of routings for the jobs A & B, both of which are due in 8 hours, is given below. Each machine is available for 8 hrs/day and no other jobs are currently scheduled from them. We will develop schedules using the forward and backward procedures.

Route Sheet job A Routine

Machine

Sequencing

Route Sheet job B Processing

Routine

Machine

Time in Hrs

Sequencing

Processing Time

in

Hrs 1

1

2

1

1

2

2

2

3

2

2

3

3

1

1

Total

6

5

The forward schedule job assigns top priority job a the earliest time slot available for the first machine on its routing i.e. first machine 1, for two hours. Then job A is assigned for 3 hours through 5 on machine 2, Job A finishes on machine 1 at the end of hour 6, its earlier possible finish time. Next job B is assigned the remaining available time slots as early as possible. First it goes to machine 1 for hours 3 & 4. Then to machine 2 for hours 6 through 8.

Forward Scheduling Machine A 8

Machine B

Machine A

Machine B

A

B

B

7 6

Backward Scheduling

A A

5

A B

4 B

A

A B

3 2 1 0 Forward and Backward scheduling for ZXC company. Backward scheduling assigns the highest priority job so that it finishes just as it is due for the next station. Therefore Job A is assigned to the last machine on its routing, machine 1, for hours 8. Then it is assigned to the preceding machine on its routing machine2, for hours 5 through 7, again as late as possible. Similarly job A is assigned to the first machine on its routing, machine 1, for hours 3 & 4. Now job B is assigned to the last machine on its routing, machine 2, for hours 8, 4 & 3. It is assigned to machine 1 for hours 1 & 2. So you see how the different systems operate ad give results for the two machines. What advantages and disadvantages do you see here in the example? 1. Both the jobs are completed in time. 2. Job A completes two hours in advance than due time 3. Job B finishes at the first machine in its routing one hour before it can go to the second on its routing. 4. In the backward schedule, job B is interrupted at machine 2 in hour 4 to allow the high priority job a to pass through. Then the job B resumes in hour 8. Let us pick up the remaining part of today’s discussion, NOW. Please pay attention. PR1ORITY SEQUENCING When jobs compete for a work center's capacity, which job should be done next? Priority

sequencing rules are applied to all jobs waiting in the queue. Then, when the work center becomes open for a new job, the one with the highest priority is assigned. CHOOSING THE RIGHT SEOUENCING RULE Many different sequencing rules are available, as we'll soon see, and the logical questions are "Which rule should I use?" "What difference does it make?" Your choice is important because a sequencing rule that performs well according to one criterion on one dimension, say minimizing inventories, may not do so well according to another, say minimizing setup costs. Some major criteria are the following: Setup costs In-process inventory costs Idle time Number or percent of jobs that are late Average time jobs are late Standard deviation of time jobs are late Average number of jobs waiting in the queue Average time to complete a job Standard deviation of time to complete a job Three of the criteria (setup costs, inventory costs, and idle time) are primarily concerned with internal facility efficiency. The more these are minimized without jeopardizing service to customers, the better the use of limited resources and chances for improved profitability.

Three of the criteria (percent of jobs that are late, average time jobs are late, and variance of time jobs are late) are more customer- or service-oriented than internal. To the extent that the values of these criteria increase, service to customers deteriorates. Finally, three of the criteria (number of jobs waiting, average time to complete a job, and variance of time to complete a job) reflect both customer service and internal efficiency. It is difficult, if not impossible, to find a sequencing rule that excels on all these criteria simultaneously. Now, let us look at:SOME PRIORITY SEQUENCING RULES The following rules are representative of the many used today in manufacturing and service industries: •

First cum first served (FCFS).

As its name suggests, incoming jobs or customers are processed in their order of arrival. FCFS is commonly applied in service industries such as banks, supermarkets', etc. •

Earliest due date (EDD).

Top priority is assigned to the waiting job whose due date is earliest. EDD ignores when the jobs arrive and the time each of them takes to process. •

Shortest processing time (SPT).

The job that can be completed in the shortest time at this work center is processed next. The jobs' due dates and order of arrival are immaterial.

•

Truncated shortest processing time (TSPT).

Jobs that have waited longer than some predetermined designated truncation time are given highest priority and are processed next. If no jobs have waited that long, then the SPT rule applies. •

Least slack (LS).

Highest priority goes to the jobs whose slack time is least. Slack is calculated as the difference of the length of time remaining until the job is due, and the length of its operation time. The order of arrival is ignored. Let's examine some of these rules and illustrate how they work. We'll apply them to the five job orders waiting in the sheet metal work center of the aircraft repair facility \"1.:: discussed earlier. Customers submitted these job orders during the past week. Rather than evaluating all of the S (1 x 2 x 3 x 4 x S), or 120, different possible sequences for these five jobs, let's evaluate the sequences created according to the . FCFS and SPT rules. First-come-first-served (FCFS) Sequencing For convenience, we assume jobs arrived in alphabetical order, so that according to FCFS job A goes first, j9b B next, and so on. Customers requested their orders be completed by the due dates listed in Table . The job flow time for this sequence is the total time each job is either waiting or being processed. Job B, for example, waits 4 days while A is being processed and then takes 17 days operation time. Job B is therefore completed in 21 days, its flow time. Our FCFS sequence results in the following: 1. Total completion time. After 55 days, all jobs are completed. 2. Average flow time.

The average flow time is 31.8, calculated by summing the flow times for all jobs and dividing by the number of jobs: (4 + 21 + 3S + 44 + 55) + 5 = 31.8

Sequencing data for a first-come-first-served (FCFS) priority rule

ABCDE Waiting jobs in FCFS sequence Processing Time (in days)

Flow

Due

Time

Date

(in days)

(in days from now)

A

4

4

6

B

17

21

20

C

14

35

18

D

9

44

12

E

11

55

12

Total 55

3. Average number of jobs in the system each day. The average number of jobs flowing in the system (waiting or being processed) from the beginning o( the sequence through the time when the last job is finished is 2.89: For the first 4 days, 5 jobs are in the system; for the next 17 days, 4 jobs are in the system; for days 22 to 35, 3 jobs are in the system, and so forth. There are 55 total days for the sequence. Hence, [5(4) + 4(17) + 3(14) + 2(9) + 1(11)] 7 55 = 2.89 jobs in the system/day 4. Average job lateness. The average number of days that jobs are late is 18.6 days. The lateness of each job is the difference of its flow time and its due date. Flow time for job A, for example, is 4 days. Since its due date is 6 days, the difference is 4-6, or - 2. Job A was finished 2 days earlier than required: no lateness. Flow time for Job B is 21 days, and its due date is 20 days; this job is one day late. Similarly, lateness for jobs C, 0, and E is 17, 32, and 43 days, respectively. Average lateness is

(0 + 1 + 17 + 32 + 43) 7 5 = 18.6 days

An advantage of the FCFS rule is its simplicity; moreover, it is "fair play" from the customer's viewpoint. However, some other rules are more desirable for effective "and efficient operations. Shortest Processing Time (SPT) The SPT rule assigns highest priority to the job order whose processing time is shortest. The SPT rule yields the data in Table and the following performance by using the sequence A, D, E, C, B: 1. Total completion time. After 55 days all jobs are completed. 2. Average flow time. The sum of flow times is (4 + 13 + 24 + 38 + 55) = 134. Average flow time is 134 7 5 = 26.8 days. 3. Average number of jobs in the system each day. Over the entire span of S5 days, S jobs are flowing in the system for 4 days while job A is being processed; 4 jobs are in the system while job D is being processed for 9 days, and so on. Thus, the average number of jobs in the system each day IS: [5(4) + 4(9) + 3(11) + 2(14) + 1(17)] 7' 55 = 2.44 jobs TABLE Sequencing data for a shortest processing time (SPT) priority Waiting Job

Processing Time in Flow Time in days

Due Date in days

(in SPT sequence)

days

from now

A

4

4

6

D

9

13

12

E

11

24

12

C

14

38

18

B

17

55

20

4.Average job lateness. The lateness for jobs in this sequence are 0,1,12, 20, and 35 days, respectively. Average lateness is: (0 + 1 + 12 + 20 + 35) 75= 13.6 days With that, we have come to the end of today’s discussions. I hope it has been an enriching and satisfying experience. See you around in the next lecture. Take care. Bye.