1.

INTRODUCTION
1.1 What is manufacturing?
Planning scheduling in manufacturing is a large research area involving many
aspects and variables on different planning levels. To understand the planning and
scheduling systems in manufacturing firstly it should be understood that how exactly in
manufacturing system tasks are done.
In a manufacturing model, a resource is usually referred to as a machine; a task that
has to be done on a machine is typically referred to as a job.
When orders that should meet customer demands are taken it should be known that
whether required raw materials are available or not to translate these orders to the jobs. If
there is no raw material available the process of getting them is also a part of the
manufacturing. Required number of raw materials and time of getting them are important
steps that affect finished goods and their production time directly. Then manufacturing orders
are converted to jobs with associated due dates. These jobs have to be processed
sometimes just on the machines and sometimes by workers in a given order or sequence.
The point is that because each type of jobs can differentiates, sequence of these machines
can be arranged again so that jobs should be performed to meet the order. In a production
process, a job may be a single operation or a collection of operations that have to be done
on various different machines.

1.2 Planning: General Definition in Manufacturing

Why is planning so important in production and manufacturing areas? In a
manufacturing duration almost in every time a lot of various decisions are made and these
decisions affect the every steps of any manufacturing process. Decision-making processes
play an important role in procurement and production, in transportation and distribution and
in information processing and communication. As the importance of these decisions
increases, then the well preparation level of these decisions are increased. But in
manufacturing systems every steps even they seem not to be important should be arranged
carefully because every small steps can affect the next other important steps. They comprise
the rather simple question Which job has to be scheduled next on a respective machine?
as well as the very serious task whether to open or close a factory.
Planning can be described as the administrative process that takes place within a
manufacturing business and which involves making sure that sufficient raw materials, staff
and other necessary items are procured and ready to create finished products according to
the schedule specified. A typical large manufacturing business engaging in production
planning will aim to maximize profitability while maintaining a satisfied consumer base. Also
planning supports decision-making by identifying alternatives of future activities and selecting
some good ones or even the best one.

Figure 1: Flow Diagram in Manufacturing System

1.3 Scheduling: General Definition in Manufacturing

Scheduling is an important tool for manufacturing and engineering, where it can have
a major impact on the productivity of a process. In manufacturing, the purpose of scheduling
is to minimize the production time and costs, by telling a production facility when to make,
with which staff, and on which equipment. Production scheduling aims to maximize the
efficiency of the operation and reduce costs.
Companies use backward and forward scheduling to allocate plant and machinery
resources, plan human resources, plan production processes and purchase materials;
- Forward scheduling is planning the tasks from the date resources become available
to determine the shipping date or the due date.
- Backward scheduling is planning the tasks from the due date or required-by date to
determine the start date and/or any changes in capacity required.
There are two key definitions in scheduling which are called offline and online
scheduling. In a problem when an offline scheduling is used, all data information about a
problem is known even the future knowledge (this is not an exact information because future

demand is uncertain) and with this feature it approximates to planning except for using of
time horizon. In online scheduling information comes from past events and current
knowledge, not from the future demand expectations. And with this information when a
decision is made it cannot be changed further. It misses the information about the new data
which is further entered to the system while current scheduled is being determined.

2. PLANNING SCHEDULING
2.1 Introduction
Planning and scheduling are forms of decision-making that are used on a regular
basis in many manufacturing industries. The planning and scheduling functions in a company
could rely on mathematical techniques and heuristic methods that doing the activities with
limited resources to the activities to be done. This allocation of resources has to be done in
such a way that the company optimizes its objectives and achieves its goals. Resources may
be machines in a workshop, runways at an airport, crews at a construction site, or processing
units in a computing environment. Activities may be operations in a workshop, take-offs and
landings at an airport, stages in a construction project, or computer programs that have to be
executed. Each activity may have a priority level, an earliest possible starting time and/or a
due date. Objectives can take many different forms, such as minimizing the time to complete
all activities, minimizing the number of activities that are completed after the committed due
dates, and so on.
There are several situation which is needed to be solved by using best planning and
scheduling methods. For example; when a job which should be performed on a certain
machine cannot be achieved because that machine is busy because of performing another
job, there will occur a delay. If there are the certain jobs that is given precedence as to be
first, planning should be done according to this preemption. Another unexpected events on
the shop floor, such as machine breakdowns or longer-than-expected processing times, also
have to be taken into account, since they may have a major impact on the schedules.
It is common practice that, based on these operational plans, detailed schedules are
worked out which define the precise timing and sequencing of individual operations as well
as the assignment of the required resources over time.
http://wps.prenhall.com/wps/media/objects/7117/7288732/krm9e_SuppJ.pdf

2.2 Scheduling Models

Manufacturing systems can differentiates by a lot of specific factors such as the
number of resources or machines, their characteristics and configuration, the level of
automation, the type of material handling system, etc. Because all these characteristics
imposed on focusing different specific features planning and scheduling models should be
determined according to the type of the manufacturing system.
The benefits of production scheduling include:

Process change-over reduction

Inventory reduction, leveling

Reduced scheduling effort

Increased production efficiency

Labor load leveling

Accurate delivery date quotes

Real time information

Main five classes of scheduling are; Project Planning and Scheduling, Machine and
Job Scheduling, Flexible Assembly Scheduling, Lot Scheduling and Supply Chain Planning
and Scheduling.

2.2.1 Project Planning and Scheduling

Project planning and scheduling is important whenever a large project, that consists
of many stages, has to be carried out. A project, such as the construction of an aircraft carrier
or a skyscraper, typically consists of a number of activities or jobs that may be subject to
precedence constraints. A job that is subject to precedence constraints cannot be started
until certain other jobs have been completed. In project scheduling,it is often assumed that
there are an unlimited number of machines or resources, so that a job can start as soon as
all its predecessors have been completed. The objective is to minimize the completion time
of the last job, commonly referred to as the makespan. It is also important to find the set of
jobs that determines the makespan, as these jobs are critical and cannot be delayed without
delaying the completion of the entire project. Project scheduling models are also important in
the planning and scheduling of services. Consider, for example, the planning and scheduling
of a large consulting project.
Another more general version of the basic project planning and scheduling problem
assumes that a jobs processing requires, besides a machine also various operators. The
workforce may consist of several different pools of operators; each pool has a fixed number
of operators with a specific skill. Because of workforce limitations, it may sometimes occur
that two jobs cannot be processed at the same time,even though both are allowed to start as
far as the precedence constraints are concerned. This type of problem is in what follows
referred to as project scheduling with workforce constraints. These types of planning and
scheduling problems often occur in practice when large projects have to be undertaken.
Examples of such projects are real estate developments,construction of power generation
centers, software developments, and launchings of spacecraft.
While its easier to find optimal solutions with little computatoinal effort for basic
project planning and scheduling, for project scheduling with workforce constraints its very
hard.
2.2.1.1 Application of Project Planning Scheduling
The precedence contsraints between jobs are the basic constraints of the project
scheduling problem. There are two types of format to show the precedence relationship. First
one is job-on-arc format which satisfies the arcs in the precedence graph represent the
jobs and the nodes represent the milestones or epochs. In the job-on-node format, the nodes
in the precedence graph represent the jobs,and the connecting arcs the precedence
relationships between the jobs.

a) Job-on-arc format

b) Job-on-node format
Figure 2: Formats of precedence graphs
First algorithm is forward procedure which finds the earliest possible completion
time.Let Cj is the earliest possible completion time of job j and Sj the earliest possible
starting time. Then, Cj= Sj + pj. Let the set {all k j} denote all jobs that are predecessors of
job j. This implies that if job k is a predecessor of job j, job k has to be completed before job j
can be started.
Second algorithm is backward procedure used to determine the latest possible
starting times and completion times of all the jobs, assuming the makespan is kept at its
minimum. The algorithm uses the Cmax,whic h is an output of the forward procedure, as an
input. In order to describe this algorithm some additional notation is needed.
Let Cj is the latest possible completion time of job j and Sj is the latest possible
starting time of job j. Sj= Cj pj. Let the set {j all k} denote all jobs that are successors of
job j.
A job of which the earliest starting time is earlier than the latest starting time is
referred to as a slack job. The difference between a jobs latest possible starting time and
earliest possible starting time is the amount of slack. A job of which the earliest starting time
is equal to the latest starting time is referred to as a critical job. A critical path is a chain of
non-slack jobs, beginning with a job that starts at time zero and ending with a job that
completes its processing at Cmax.
Example: Consider 14 jobs. The processing times are given below.

Table 1: Processing times of jobs

Precedence constraints are in Figure 3.

Figure 3: Precedence graph

The earliest completion time Cj of job j can be computed using the forward procedure.

Table 2: Earliest Completion Time

Makespan is 56. Assuming that the makespan is 56, the latest possible completion
times Cj can be computed using the backward procedure.

Table 3: Latest Completion Time

Those jobs of which the earliest possible completion times are equal to the latest
possible completion times are critical and constitute the critical path. So the critical path is

The critical path in this case happens to be unique. The jobs that are not on the
critical path are slack.
According to the CP rule, whenever a machine is freed,the job at the head of the
longest chain is given the highest priority. Similarly, in the CPM method one has to make sure
that the job at the head of the longest chain is never delayed. However,since there are an
unlimited number of machines and no workforce constraints in the environment considered
here, other jobs do not have to be delayed either. The method, does not function as a priority
rule. Generally, with a limited number of machines or with workforce constraints, the Critical
Path rule may actually have to postpone the start of jobs that are at the head of shorter
chains.

2.2.2 Machine and Job Scheduling

The second class of models include single machine, parallel machine and job shop
models. In a single machine or parallel machine environment, a job consists of one operation
that can be done on any one of the machines available. In a full-fledged job shop, a job
typically consists of a number of operations that have to be performed on different machines.
Each job has its own route that it has to follow through the system. The operations of the jobs
in a job shop have to be scheduled to minimize one or more objectives, such as the
makespan or the number of late jobs. Job shops are prevalent in industries that make
customized industrial hardware. However,they also appear in service industries (e.g.,ho
spitals). A special case of a job shop is a setting where each one of the jobs has to follow the
same route through the system (i.e., each job has to be processed first on machine 1, then
on machine 2, and so on); such a setting is usually called a flow shop.
2.2.2.1 Single Machine and Parallel Machine Models
Consider a single machine and n jobs. Job j has a processing time pj, a release date
rj and a due date dj. If rj = 0 and dj = , then the processing of job j is basically unconstrained.

It is clear that the makespan Cmax in a single machine environment does not depend on the
schedule. For various other objectives certain priority rules generate optimal schedules.
If the objective to be minimized is the total weighted completion time, i.e., wj Cj, and
the processing of the jobs is unconstrained, then the Weighted Shortest Processing Time
first (WSPT) rule, which schedules the jobs in decreasing order of wj/pj, is optimal. If the
objective is the maximum lateness Lmax and the jobs are all released at time 0, then the
Earliest Due Date first (EDD) rule, which schedules the jobs in increasing order of dj, is
optimal. Both the WSPT rule and the EDD rule are examples of static priority rules. A rule
that is somewhat related to the EDD rule is the so-called Minimum Slack first (MS) rule which
selects, when the machine becomes available at time t, the job with the least slack; the slack
of job j at time t is defined as max(dj pj t, 0). This rule does not operate in exactly the same
way as the EDD rule,but will result in schedules that are somewhat similar to EDD
schedules. However,the MS rule is an example of a dynamic priority rule,in which the priority
of each job is a function of time.
2.2.2.2 Job Shops
Consider a job shop with n jobs and m machines. Each job has to be processed by a
number of machines in a given order and there is no recirculation. The processing of job j on
machine i is referred to as operation (i, j) and its duration is pij . The objective is to minimize
the makespan Cmax. Solution technique of job shop model with an example will be
discussed later.

2.3 Flexible Assembly Scheduling

The third class of models focuses on production systems with automated material
handling. In these settings a job also consists of a number of operations. A material handling
or conveyor system controls the movement of the jobs as well as the timing of their
processing on the various machines. Examples of such environments are flexible
manufacturing systems, flexible assembly systems,a nd paced assembly lines. The objective
is typically to maximize throughput. Such settings are prevalent in the automotive industry
and in the consumer electronics industry.

2.4 Lot Scheduling

The fourth class of models are known as lot scheduling models. These models are
used for medium and long term production planning. In contrast to the first three classes, the
production and demand processes are now continuous. In this class, there are a variety of
different products. When a machine switches from one product to another, a changeover cost
is incurred. The goal is usually to minimize total inventory and changeover costs. These
models are important in the process industries, such as oil refineries and paper mills.

2.5 Supply Chain Planning and Scheduling

The fifth class of models consists of supply chain planning and scheduling models.
These models tend to be hierarchical and are often based on an integration of the lot
scheduling models (the fourth class of models) and the job shop scheduling models (the
second class of models). The objective functions in supply chain planning and scheduling
take into account inventory holding costs at the various stages in the chain as well as costs
of transportation between the stages. There are restrictions and constraints on the production
quantities as well as on the quantities that have to be transported from one stage to another.

3. TECHIQUES USED IN PLANNING SCHEDULING

3.1 Introduction
There are many solution techniques available to use the resources such as labor,
money, time, raw material or machinery most effectively while making a plan in production
which means producing most number of products from minimum number of resources. The
aim is that finding an optimum solution to the problems to increase productivity, get maximum
profit and have minimum cost. Sometimes it is not possible to reach the optimal solution,
then instead of this what the nearest optimal solution should be determined. And also
according to the type of the problem, suitable solution technique should be determined.
Three of the solution techniques are Linear Programming, Non-Linear Programming
and Shifting Bottle Neck Heuristic.

3.2 Linear Programming

Linear Programming (LP) is a mathematical technique designed to help operations
managers plan and make decisions relative to the trade-offs necessary to allocate resources.
Many planning and scheduling decisions involve trying to make the most effective use of an
organizations resources. Machinery (such as planes, in the case of an airline), labor (such
as pilots), money, time, and raw materials (such as jet fuel) are described as resources.
These resources may be used to produce products (such as machines, furniture, food, or
clothing) or services (such as airline schedules, advertising policies, or investment
decisions).
As a summary two of classic fields of linear programming used;
1. Manufacturing --product choice
-Several alternative outputs with different input requirements,
-Scarce inputs,
-Maximize profits.
2. Scheduling
-Many possible personnel shifts,
-Staffing requirements at various times,
-Restrictions on shift timing and length,
-Minimize cost of meeting staffing requirements.
A few examples of problems in which LP has been successfully applied in
production/manufacturing are;
-Selecting the product mix in a factory to make best use of machine and labor-hours
available while maximizing the firms profit.
-Developing a production schedule that will satisfy future demands for a firms product
and at the same time minimize total production and inventory costs.
All LP problems have four properties in common:

1. LP problems aim to maximize or minimize some quantity (usually profit or cost but
it can be any goal). This is called as objective function of an LP problem. The major
objective of a typical firm is to maximize profits in the long run. In the case of a trucking or
airline distribution system, the objective might be to minimize shipping costs.
2. In LP problems there are always constraints that are restrictions limiting the
degree to which we can pursue our objective. We should decide the optimum solution which
is maximizing or minimizing quantity (objective function) according to our limited resources
(constraints). For example, deciding how many units of each product in a firms product line
to manufacture is restricted by available labor and machinery.
3. There must be alternative courses of action to choose from. For example, if a
company produces three different products, management may use LP to decide how to
allocate among them its limited production resources (of labor, machinery, and so on). If
there were no alternatives to select from, we would not need LP.
4. The objective and constraints in linear programming problems must be expressed
in terms of linear equations or inequalities.

3.2.1 Linear Programming Model

Let:

X1, X2, X3, , Xn = decision variables

Z = Objective function or linear function

Objective: Maximization or minimization of the objective function Z;
Z = c1 X1 + c2 X2 + c3 X3 + +cn Xn

Equation (1)

Subject to the following constraints;

a11 X1 + a12 X2 + a13 X3 .. +a1n Xn b1
a21 X1 + a22 X2 + a23 X3 .. +a2n Xn b2
.

Equation (2)

am1 X1 + am2 X2 + am3 X3 .. +amn Xn bm

all Xn 0
Where an, bn, cn are given constraints.
WORK-SCHEDULING EXAMPLE
A post office requires different numbers of full-time employees on different days of the
week. The number of full-time employees required on each day is given in Table 4. Union
rules state that each full-time employee must work five consecutive days and then receive
two days off. For example, an employee who works Monday to Friday must be off on
Saturday and Sunday. The post office wants to meet its daily requirements using only fulltime
employees. Formulate an LP that the post office can use to minimize the number of full-time
employees who must be hired.

In Linear Programming the first important thing is defining the decision variables correctly.
Here we are defining the decision variables as;
xi = number of employees beginning the work on day i.
For example, x1 is the number of employees beginning work on Monday and finish the work
on Friday. After determining the variables properly, its easy to determine objective function
and constraints. The objection function should be minimizing total number of employees;
min z = x1 + x2 + x3 + x4 + x5 + x6 + x7
The next step is determining the constraints. First constraint concerning Monday is number
of employees working on Monday should be at least 17. It should be noticed that, for
example, employees starting work on Tuesday and Wednesday dont work on Monday. The
constraint of Monday is;
x1 + x4 + x5 + x6 + x7 17
When other constraints of other days are determined by this way, linear programming
formulation of this problem is;
min z = x1 + x2 + x3 + x4 + x5 + x6 + x7
x1 +
x1 + x2 +

x4 + x5 + x6 + x7 17

(Monday constraint)

+ x5 + x6 + x7 13

(Tuesday constraint)

x1 + x2 + x3 +

+ x6 + x7 15

x1 + x2 + x3 + x4 +

+ x7 19

x1 + x2 + x3 + x4 + x5 +
+ x 2 + x 3 + x4 + x 5 + x 6 +

(Thursday constraint)

14

(Friday constraint)

16

(Saturday constraint)

+ x3 + x4 + x5 + x6 + x7 11
xi 0

(Wednesday constraint)

(i = 1, 2, . . . , 7)

(Sunday constraint)
(Sign restrictions)

Modelling Issues: This model is an example of static scheduling problem. Its

assumed that in the post office each week same schedule is used. However, in real life the
demands in the post office is changable and employees might have vacations. This example
is just for explaning the model of Linear Programming.

3.2.2 Solution Techniques of Linear Programming

3.3 Nonlinear Programming

In a linear programming, the objective function and constrait equations are linear
functions in problems. But, in other problems objective function or constraints may not be a
linear equation. These problems are called as nonlinear programming problems and there
are many techniques to solve these problems.

3.3.1 Nonlinear Programming Model

x1, x2, x3, , xn = decision variables
z = Objective function or linear function
Objective: Maximization or minimization of the objective function z;
min ( or max )

z = f(x1, x2, x3,.....,xn)

subject to following constraints:

g1(x1, x2, x3,.....,xn) (, or =) b1
g2(x1, x2, x3,.....,xn) (, or =) b2
.
.
gm(x1, x2, x3,.....,xn) (, or =) bm
If any Nonlinear Programming Problems have no constraints, they are called as
unconstrained NLP.

3.3.2 Solution of NLPs

Feasible Points
The feasible region for NLP is the set of points (x1, x2,...,xn ) that satisfy the m
constraints in (. A point in the feasible region is a feasible point, and a point that is not in the
feasible region is an infeasible point.
Optimal Solutions
If problem is a maximization problem, any point xxin the feasible region for which f( xx
) f(x)
holds for all points x in the feasible region is an optimal solution to the NLP.
For a minimization problem, xxis the optimal solution if f( xx
) f(x) for all feasible x.
Of course, if f, g1 ,g2 ,..., gm are all linear functions, then it is a linear programming problem
and may be solved by the simplex algorithm.

3.4 Shifting Bottleneck Heuristic

Since it is very hard to solve job shop problems with large numbers of jobs to find
optimal solution, many heuristic procedures have been designed. One of the most successful
procedures for minimizing the makespan in a job shop is the Shifting Bottleneck heuristic.
In the following overview of the Shifting Bottleneck heuristic M denotes the set of all m
machines. In the description of an iteration of the heuristic it is assumed that in previous

iterations a selection of disjunctive arcs has been fixed for a subset M0 of machines. So for
each one of the machines in M0 a sequence of operations has already been determined.
An iteration results in a selection of a machine from M M0 for inclusion in set M0.
The sequence in which the operations on this machine are to be processed is also generated
in this iteration. To determine which machine should be included next in M0,an attempt is
made to determine which unscheduled machine causes in one sense or another the severest
disruption. To determine this, the original directed graph is modified by deleting all disjunctive
arcs of the machines still to be scheduled (i.e.,the machines in set MM0) and keeping only
the relevant disjunctive arcs of the machines in set M0 (one from every pair). Call this graph
G. Deleting all disjunctive arcs of a specific machine implies that all operations on this
machine,whic h originally were supposed to be done on this machine one after another,no w
may be done in parallel (as if the machine has infinite capacity,o r equivalently, eachone of
these operations has the machine for itself). The graph G has one or more critical paths that
determine the corresponding makespan. Call this makespan Cmax(M0).
Suppose that operation (i, j), i {MM0},has to be processed in a time window of
which the release date and due date are determined by the critical (longest) paths in G ,i.e.,
the release date is equal to the longest path in G from the source to node (i, j) and the due
date is equal to Cmax(M0),minus the longest path from node (i, j) to the sink, plus pij .
Consider each of the machines in M M0 as a separate nonpreemptive single machine
problem with release dates and due dates and with the maximum lateness to be minimized.,
this problem is NP-hard; however, procedures have been developed that perform reasonably
well. The minimum Lmax of the single machine problem corresponding to machine i is denoted
by Lmax(i) and is a measure of the criticality of machine i.
After solving all these single machine problems, the machine with the largest
maximum lateness Lmax(i) is selected. Among the remaining machines, this machine is in a
sense the most critical or the bottleneck and therefore the one to be included next in M0.
Assume this is machine h, call its maximum lateness Lmax(h) and schedule it according to the
optimal solution obtained for the single machine problem associated with this machine. If the
disjunctive arcs that specify the sequence of operations on machine h are inserted in graph
G,then the makespan of the current partial schedule increases by at least Lmax(h),tha t is,
Cmax(M0 h) Cmax(M0) + Lmax(h).
Before starting the next iteration and determining the next machine to be scheduled, a
n additional step has to be done within the current iteration. In this additional step all the
machines in the original set M0 are resequenced one by one in order to see if the makespan
can be reduced. That is, a machine, say machine l, is taken out of set M0 and a graph G is
constructed by modifying graph G through the inclusion of the disjunctive arcs that specify
the sequence of operations on machine h and the exclusion of the disjunctive arcs
associated with machine l. Machine l is resequenced by solving the corresponding single
machine maximum lateness problem with the release and due dates determined by the
critical paths in graph G. Resequencing each of the machines in the original set M0
completes the iteration.
In the next iteration the entire procedure is repeated and another machine is added to
the current set M0 h. The shifting bottleneck heuristic can be summarized as follows.
Algorithm of Shifting Bottleneck Heuristic
Step 1. (Initial conditions)

Set M0 = .
Graph G is the graph with all the conjunctive arcs and no disjunctive arcs.
Set Cmax(M0) equal to the longest path in graph G.
Step 2. (Analysis of machines still to be scheduled)
Do for each machine i in set M M0 the following: formulate a single
machine problem with all operations subject to release dates and due dates
(the release date of operation (i, j) is determined by the longest path in
graph G from the source to node (i, j); the due date of operation (i, j) can
be computed by considering the longest path in graph G from node (i, j) to
the sink and subtracting pij ).
Minimize the Lmax in each one of these single machine subproblems.
Let Lmax(i) denote the minimum Lmax in the subproblem corresponding to
machine i.
Step 3. (Bottleneck selection and sequencing)
Let Lmax(h) = maxi{MM0}(Lmax(i))
Sequence machine h according to the sequence generated for it in Step 2.
Insert all the corresponding disjunctive arcs in graph G.
Insert machine h in M0.
Step 4. (Resequencing of all machines scheduled earlier)
Do for each machine l {M0 h} the following:
Delete the corresponding disjunctive arcs from G; formulate a single machine
subproblem for machine l with release dates and due dates of the
operations determined by longest path calculations in G.
Find the sequence that minimizes Lmax(l) and insert the corresponding
disjunctive arcs in graph G.
Step 5. (Stopping criterion)
If M0 = M then STOP, otherwise go to Step 2.
The structure of the shifting bottleneck heuristic shows the relationship
between the bottleneck concept and the more combinatorial concepts such as critical
(longest) path and maximum lateness. A critical path indicates the location and the
timing of a bottleneck. The maximum lateness gives an indication of the amount by
which the makespan increases if a machine is added to the set of machines already
scheduled. The following example illustrates the use of the shifting bottleneck
heuristic.

Example of application of Shifting Bottleneck Heuristic

Consider the instance with four machines and three jobs. The routing,i .e., the
machine sequences, and the processing times are given in the following table:
jobs machine sequence processing times:

Iteration 1: Initially, set M0 is empty and graph G contains only conjunctive arcs and
no disjunctive arcs. The critical path and the makespan Cmax() can be determined easily:
this makespan is equal to the maximum total processing time required for any job. The
maximum of 22 is achieved in this case by both job 1 and job 2. To determine which machine
to schedule first, eachmachine is considered as a nonpreemptive single machine maximum
lateness problem with the release dates and due dates determined by the longest paths in G
(assuming a makespan of 22).

Figure 2. Iteration 1 of shifting bottleneck heuristic

The data for the nonpreemptive single machine maximum lateness problem corresponding to
machine 1 is determined as follows:

The optimal sequence turns out to be 1, 2, 3 with Lmax(1) = 5.

The data for the subproblem associated with machine 2 are:

The optimal sequence for this problem is 2, 3,1 with Lmax(2) = 5. Similarly, it can be shown
that:
Lmax(3) = 4
and

Lmax(4) = 0.
From this it follows that either machine 1 or machine 2 may be considered a bottleneck.
Breaking the tie arbitrarily, machine 1 is selected to be included in M0. The graph G is
obtained by fixing the disjunctive arcs corresponding to the sequence of the jobs on machine
1 (see Figure 2). It is clear that
Cmax({1}) = Cmax() + Lmax(1) = 22 + 5 = 27.
Iteration 2: Given that the makespan corresponding to G is 27,the critical paths in the graph
can be determined. The three remaining machines have tobe analyzed separately as
nonpreemptive single machine problems. The data for the problem concerning machine 2
are:

The optimal schedule is 2, 1, 3 and the resulting Lmax(2) = 1. The data for the problem
corresponding to machine 3 are:

Both sequences are optimal and Lmax(3) = 1. Machine 4 can be analyzed in the same way
and the resulting Lmax(4) = 0. Again, there is a tie and machine 2 is selected to be included
in M0. So M0 = {1, 2} and
Cmax({1, 2}) = Cmax({1}) + Lmax(2) = 27 + 1 = 28.
The disjunctive arcs corresponding to the job sequence on machine 2 are added to G and
graph G is obtained. At this point,st ill as a part of iteration 2,an attempt may be made to
decrease Cmax({1, 2}) by resequencing machine 1. It can be checked that resequencing
machine 1 does not give any improvement.
Iteration 3: The critical path in G can be determined and machines 3 and 4 remain to be
analyzed. These two problems turn out to be very simple with both having a zero maximum
lateness. Neither of the machines constitutes a bottleneck in any way.
The final schedule is determined by the following machine sequences: the job sequence 1, 2,
3 on machine 1; the job sequence 2, 1, 3 on machine 2; the job sequence 2, 1 on machine 3
and the job sequence 2, 3 on machine 4. The makespan is 28.

4. TYPES OF PLANNING
Planning is one of the four major functions of management. To achieve the
organizational goal, it is important to understand the planning concept exactly.
There are three main types planning of planning. These are strategic, tactical and
operatinal planning. If the relationship between these different planning stages are connected
each other tightly, it will be easier to get succesful consequiences in manufacturing systems.
Because one of these types is affected by results of the other type of planning. Operational
plans are necessary to attain tactical plans and tactical plans lead to the achievement of
strategic plans.

4.1 Strategic Planning

To better understand the relationship between the three types of planning, we should
start with the top. Strategic plans are designed with the entire organization in mind and begin
with an organization's mission. Top-level managers, such as CEOs or presidents, will design
and execute strategic plans to paint a picture of the desired future and long-term goals of the
organization. Essentially, strategic plans look ahead to where the organization wants to be in
three, five, even ten years. Strategic plans, provided by top-level managers, serve as the
framework for lower-level planning.
Strategic plans can be for achieving growth, improving productivity and profitability
and boosting return on investments which are all part of the desired future of the
manufacturing system. Strategic plans also tend to require multilevel involvement so that
each level of the organization plays a significant role in achieving the goals being
strategically planned for. Top-level managers, develop the organizational objectives so that
middle- and lower-level managers can create compatible plans aligned with those objectives.

4.2 Tactical Planning

Second level of planning, known as tactical planning. Tactical plans support strategic
plans by translating them into specific plans relevant to a distinct area of the organization.
Tactical plans are concerned with the responsibility and functionality of lower-level
departments to fulfill their parts of the strategic plan.
When strategic plans are completed, middle level manager begins to make tactical
plans to achieve determined strategic plans. Middle-level managers who are responsible for
tactical planning, manage the operations which should be arranged as increasing productivity
of manufacturing. He should create a set of calculated actions that satisfy the better
processes which achieve organizational goal and are narrower in scope than the strategic
plan but still help to bring the organization closer to the long-term goal.

4.3 Operational Planning

Operational plans are at the bottom of the whole planning; they are the plans that are made
by frontline, or low-level, managers. All operational plans are focused on the specific
procedures and processes that occur within the lowest levels of the organization. Managers
must plan the routine tasks of the department using a high level of detail.
Operational planning activities for lower-level manager would include things like
scheduling employees each week; assessing, ordering and stocking inventory; creating a

monthly budget; developing a promotional advertisement for the quarter to increase the sales
of a certain product or outlining an employee's performance goals for the year.
Operational plans can be either single-use or ongoing plans. Single-use plans are
those plans that are intended to be used only once. They include activities that would not be
repeated and often have an expiration. Ongoing plans are those plans that are built to
withstand the test of time. They are created with the intent to be used several times and
undergo changes when necessary. Outlining an employee's performance goals for the year
would be considered an ongoing plan, assess and update, if necessary.

5. ADVANCED PLANNING AND SCHEDULING

There is an historical development which trying to satisfy the optimal planning
scheduling in manufacturing. Defining these developments better explains the importance of
Advanced Planning and Scheduling.
Before 1950, planning was done manually. Computers were not used for planning
issues. In 1960s Material Requirements Planning (MRP) was developed with Operations
Research techniques. Then, in 1970s its extended to MRP II which means Manufacturing
Resource Planning; provide users checking the feasibility of capacity with feedback loops at
various stages. This turned the planning issue from theory into practice. In 1980, ERP is
introduced. ERP reached the wider scope in manufacturing such as finance, purchasing, HR
and marketing. To understand the APS concept, it would be better to compare it with ERP.
MRP makes planning with firstly calculating material requirements and assuming that
having infinite capacity. And then required capacity is calculated. If the MRP outcome does
not fit the limited capacity, which means plan is not feasible, you have to make adjustments
and start over again. The complexity here is that it does not tell you which changes you
should make. To make the plan fit, which products you should produce over others?
However, APS plans material requirement and capacity simultaneously and this guarantee
you always getting the feasible planning. Even more, planning is optimal instead of being
feasible because within the capacity constraints, APS determines for you what optimal
planning is. In short: with using MRP, you know what is going on, but with APS you know
what should be going on.
Because APS systems are capable of running a plan much faster compared to MRP,
they are also much better suited to adapt to changes in demand, resource capacity or
material availability. It even allows a planning department to simulate different scenarios.
Typically, an APS system is capable of planning over multiple stages in a supply chain and
will support planning on multiple levels (strategic, tactical and operational). In table 1, with
the differences, application field and advantages of APS.

Table 3: Differences between ERP and APS

APS system
Customer preference may be varied
depending on the business importance of
the customer
Lead times can be dynamically entered by
contacting the customers
APS applications dynamically calculate a
plan and schedule within minutes of any
change being made to them
Support superior decision making by whatif
analysis and simulations
Smart and easy to drill down reporting
based on the identification of exceptional
conditions
Material allocation according to availability
and according to the criterion specified

MRP/II system
All customers are given equal preference
in the system
Lead times are fixed and known a priori
MRP runs are usually batch time and have
longer duration times
Does not support any decision making aids
Detailed reports, which are hard to read
and decipher
Material allocation done on a first come
first service basis

Table 4: Comparison APS systems and MRP/II systems

APS system
Facilitate real time analysis for planning,
scheduling and optimization decision

Material and capacity constraints are

evaluated together to arrive at an optimal
decision
Multi-plant planning is supported
Lead times can be calculated dynamically
Production schedules can be optimized to
increase throughput
Results can be entered into the system to
prove the processes and data

ERP system
Level of detail is coarser and the
technology does not support real time
analysis
and simulation to aid dynamic decision
making
No consideration for interdependency of
material and capacity availability
Multi-plant planning not supported at the
same time
Lead times are assigned statistically and
manually
Absence of optimization capability for
production schedules
Results can be entered into the system to
improve the process and data

Table 5: Comparison of APS systems and ERP systems

5.1 Elements of an Advanced Planning and Scheduling System (APS)

APS is a tool designed to help manufacturers develop an attainable schedule while
balancing internal constraints and limited resources. APS balances due dates, machine
capacity, tooling and labor to develop a realistic plan of action to move orders through
various operation steps. Considerable modeling effort, application maintenance and
database integrity is necessary. Benefit provided users is that generated schedules are
realistic and achievable on the plant floor because the production constraints on the plant
floor are considered and modeled.

Track each order through each respective routing step and know where each part
should be at any particular time.

Keep track of all required resources for each manufacturing step and know the
availability of each resource at any particular time.
Represent time in a detailed manner (e.g. minute by minute). Machine run time can
be defined as pounds per hour, lot time, or time per specified number of pieces.
Schedule orders (i.e. manufacturing steps) only when all critical resources are
available.
Detailed calendars and shift patterns for operators and machines.
Ability to organize machines into work centers, production lines and plant areas
(departments).
Ability to share setup time between similar products.
User-definable scheduling rules with multiple scheduling parameters.
Scheduling of preventative maintenance and other machine down time.
Constraints of raw material and component availability.
Optimized schedule based on product characteristics or attributes.

APS capabilities

Description

Capacity modelling

Exact definition of resources

and constraints
Create routing on a
productby-product basis, set
up alternate workstations
and operations workflows
that help tie workstations
together and split operations
in to separated tasks, and
place limitations or special
setupon operations
Schedule and optimize
various jobs and process
performance criterion based
on available operators and
resource constraints
Plan resource and facilities
for the long term, through
what-if analysis support the
available to promise
quantities improve processes
and discover production
constraints, and allocate
resource to specific tasks
Identify constraints and
reschedule operators, and
resources in case of change
in priorities of demand
Manage the operations by
exceptions and through
smart
easy to drill down reporting

Route modelling

Scheduling and optimization

Planning capabilities

Constraint management and

analysis
Execution control

Example Application and

Solutions
Demand planning module,
supply chain planner, supply
chain strategic, factory
planner, transportation
modeling and analysis (i2
Technologies)

Demand planning, supply

management, network
design
and optimization (JDA)

Production planning and

detailed scheduling, global
available to promise (SAP)

Table 6: Application capabilities and solution details of APS systems. The name of
application suite is followed by the name of vendor, offering the application in parentheses.

7. SAP APO
7.1 Introduction to SAP
Consider a large enterprise like PepsiCo, which has number of divisions under it.
There is a financial department, logistics section, HR, Warehousing, Sales and distribution
etc. All these need to be integrated together, for effective functioning. This is done by a
specific software known as Enterprise Resource Planning or ERP.
Five ex IBM employees started SAP as a small software and just one customer in
Germany.. Those visionary entrepreneurs started SAP in 1972. They started it with the vision
of the software that process data when a user wants, when a customer wants and not in
overnight batch jobs like earlier software. SAP started with financial application. Modules
such as Logistics, HR were added later on.
In earlier years, SAP started as R/2 that is Real time architecture with 2 servers. This
got changed in later years as R/3 that is Real time architecture with 3 servers. These 3
servers are Application Server, Production Server and Database server. In 2009, SAP
released its latest version that is ECC 6.0.
SAP is an ERP system. ERP stands for Enterprise Resources Planning. ERP term
used for software that controls whole organizations different departments. For example, SAP,
Oracle, People soft, JD Edwards are some of the top ERP software systems.
SAP is the short form of Systems, Applications and Products in data processing. SAP
as an ERP software has five functional modules and five technical modules. The functional
modules are Finance & Control, Production Planning, Material Management, Sales &
Distribution and Human Resources. Advanced Business Application Programming, Exchange
Infrastructure, Net Viewer, Basis and Business Information Warehousing come under the
technical module.
Modules present in SAP
1. Financial Accounting
2. Financial Supply Chain Management.
3. Controlling.
4. Material Management.
5. Sales and Distribution.
6. Logistic Execution.
7. Production Planning.
8. Quality Management.
9. Plant Maintenance.
10. Project system.
11. Human Resources.
7.2 SAP APO

It consists of three basic functionalities: demand planning, inventory in distribution

network and production planning and scheduling. Each of them is connected with a so-called
alert monitor, which automatically notifies planners on emergencies, such as inventory
overloads, delays in implementation of production in relation to scheduling, material
shortages, inadequate safety stocks. From the technological side APO ensures use of
efficient algorithms and technologies enabling implementation of complex calculations in real
time.
SAP APO (Advanced Planner & Optimizer) is an APS class system (Advanced
Planning System) for advanced planning and supply chain optimization. According to SAP
literature, Advanced Planner and Optimizer is designed to help a company improve
production planning, pricing, scheduling, and product shipping. APO works by getting realtime updates from retailers about customer demand. The updates are used to create APO
"demand triggers" that take into account many complex variables, such as the delivery
schedule of raw materials and productions cycles, to forecast the right amount of product mix
the company will need to meet future customer demands. APO can be integrated with the
SAP R/3 and legacy enterprise resource planning (ERP) systems.
APO consists of eight application levels: network design, demand planning, supply
network planning, production planning and detailed scheduling, global availability,
transportation planning and vehicle scheduling, and supply chain collaboration. It consists of
three basic functionalities: demand planning, inventory in distribution network and production
planning and scheduling. Each of them is connected with a so-called alert monitor, which
automatically notifies planners on emergencies, such as inventory overloads, delays in
implementation of production in relation to scheduling, material shortages, inadequate safety
stocks. From the technological side APO ensures use of efficient algorithms and technologies
enabling implementation of complex calculations in real time.
Components of SAP APO
SAP APO is not a standalone product, but integrates with SAP ERP which supplies
master and the transaction production, sales and material data in the real time to the APO
application.
Demand Planning - The demand planning component of APO can create a forecast
of market demand for a company's products. Demand Planning is a flexible tool that
supports the demand planning process in a company. Using the demand planning
library of statistical forecasting and advanced macro techniques it is possible to
create forecasts based on demand history as well as any number of causal factors,
carry out predefined and self-defined tests on forecast models and forecast results,
and adopt a consensus-based approach to reconcile the demand plans of different
departments. Promotions and forecast overrides can be used to add marketing
intelligence and make management adjustments.
Supply Network Planning - This component integrates purchasing, manufacturing,
distribution, and transportation allowing tactical planning and sourcing decisions to be
simulated and implemented using a single consistent model. Supply Network
Planning uses advanced optimization techniques, based on constraints and penalties,
to plan product flow along the supply chain. The result is optimal purchasing,
production and distribution decisions, reduced order fulfillment times and inventory
levels, and improved customer service. This component is used to calculate
quantities to be delivered to a location in order to match customer demand and

maintain the desired service level. It includes both heuristics and mathematical
optimization methods to ensure that demand is covered and transportation,
production, and warehousing resources are operating within the specified capacities.
Production Planning and Detailed Scheduling (PP/DS) - This component allows
the planning and optimization of multi-site production while taking into account
product and capacity availability. PP/DS is designed to plan critical products, such as
those with a long replenishment lead time or those that are produced on bottleneck
resources. With the plans generated through PP/DS, it is possible to reduce lead
times, increase on-time delivery, increase throughput due to better management and
planning of resources, and reduce inventory costs.
Global Availability to Promise - This functionality is required by companies as they
manufacture and sell items in the global marketplace. A basic method of availability
check is to perform an availability check on the items required, check any existing
product allocations, and to perform a forecast check. An availability check on the
items required is performed on the available to promise (ATP) quantity. This quantity
is calculated from the current stock, planned receipts, and planned requirements. The
system dynamically checks stock and planned goods movements. The global ATP will
also check any existing product allocations which may have already been made to
specific customers or for a certain region that requires a longer shipping time. The
forecast check is important as it allows users to identify if enough planned
independent requirements are available for the incoming sales orders.
Supply Chain Cockpit
An important function of SAP APO is the Supply Chain Cockpit (SCC). It provides
users with a graphical instrument panel for managing and controlling the supply chain. It is
possible to configure the SCC so that it reflects the requirements needed of the business.
The SCC can be used by a number of different areas within a company, such as strategic
planners, demand planners, and production planners.
The SCC is divided into three areas; the upper display (which shows the current
settings, which consist of model, planning work area, planning period, and view), the tree
structure on the left of the screen, and the specific model which is shown on the right.
http://logistics.about.com/od/supplychainsoftware/a/Sap-Advanced-Planning-AndOptimizer-apo.htm
http://sidgroup.pl/en/sap-sid/sap-scmsupply-chain-management/advanced-planningand-optimization-sap-apo/

EXAMPLE OF AN APS
In order to elaborate the advanced planning and scheduling in a company, it is
necessary to create a model for the production environment of said company, so that the
planning and scheduling consider all the characteristics and restrictions of the studied
production environment. This modeling is accomplished only once and will remain the same,
while there is no change in the production environment. Following that, this modelling is
registered in the APS system and will be available to make further planning and scheduling.
Many companies already have most information needed to create a model for its
productive process registered in its management system (ERP) such as: the structure of the
manufactured products, production time, precedence list for the production operations, and
others. In case the company already has the available information, the same can be
imported by the APS system. Otherwise, it will be necessary to register them in the APS
system.
Here, a hypothetical metal-mechanic industry will be taken as a example. It
manufactures replacement pieces for all types of machines. Also, this company wishes to
use the APS system to perform the planning and scheduling to a specific piece, known in this
situation as Replacement Piece.
The first information is list of materials which are pieces. The Figure 3 presents the
structure to the product Replacement Piece. To produce this product, Part 1 and Part 2
components are necessary.

Figure 3: Replacement Piece Product Structure.

Second step is to determine the available and required resources, and work schedule
of production shift. Consider that this industry has a single Monday-through-Friday shift,
starting work at 07 a.m., lunchtime at 11 and back to work at 1 p.m., finalizing the shift at 5
p.m. Consider that all production resources human resources (labor) and also the physical
resources (machines), adhere to this single work shift. Resources are shown in Table 10.

Table 10: Production resources available.

The product Replacement Piece follows the production route according to Figure 4.

Figure 4: "Replacement Piece" production route.

Each rectangle represents a productive operation. The dotted rectangles represent all
the productive operations necessary to produce a part or component of the final product. The
arrows between the operations represent the sequence among these operations, that is, the
route the product follows to be produced. The operations interconnected by double-tipped
arrows are alternative operations, that is, only one of the alternative operations must be
used. Also, the setup and the unitary execution time for each productive operation
composing the route of the product to be manufacture were also informed.
Other step is to identify which types of resources will be necessary to perform each
one of the productive operations.

Table 2 :Replacement Piece Product Structure.

With these pieces of information, it is possible to create a model of the productive
environment in the advanced planning and scheduling system. Once the information is
obtained, the next step is to transfer them to the APS system to create the productive
environment model from this example industry. In the Figure 3 there is the interface with the
user of the example APS system.

Figure 3 : APS system user interface.

All necessary information are registered to Data. The first action to be done in the
APS system is to register the product Replacement Piece and the two metallic parts
composing it. After, it is necessary to register the work calendar of the resources, according
to the example companys work shift. Then the necessary resource groups are registered to
produce the Replacement Piece in the simulated industry. Still regarding the production
environment registers, the production resources available in each resource group must now
be registered. The next step is to register the productive operations which will be necessary
to produce each part of the manufactured products. After the registration of the productive
operations to each product, it is necessaary to inform the secondary resources, which will be
necessary in each registered operation. By doing it, all the necessary data on the production

environment are already informed, that is, a model of the productive environment was
created in the APS system to the manufacture of the product Replacement Piece to this
example industry. This model can be stored in a data file which can be appropriately saved,
and its use is allowed whenever a production scheduling is wished.
Once all the information on the production environment is registered, the structure of
the product to be scheduled must be defined. To do this, APS system makes a graphic
interface available, the specific desktop Product Structure. In addition to the structure, it is
also part of the definition process of the productive environment the definition of the
production route of the products. Therefore, the example APS system makes available a
specific desktop, in which one can view and edit graphically the production route of a
product.

Figure 4: Desktops Product Structure and Production Route.

4. Operation of the APS system in the example industry
First thing is to perform production scheduling is to view which are the orders to be
scheduled. This is done in the example APS system in the specific desktop Orders to be
Scheduled, and it is possible to view and select the production orders which will be
scheduled, as well as to alter information from these orders.
In figure 5 it is possible to view the specific desktop of orders to be scheduled. In this
desktop, it is possible to select the production orders which will be scheduled, in addition to
check the details of these orders. Here it was selected the first production order registered
(marked in blue). Then it was performed the scheduling of the selected order, as shown in
figure 5. After the production scheduling is done, the APS system informs the real delivery
date scheduled.

Figure 5 : Desktop Orders to be Scheduled with the orders to be scheduled and the
delivery date of a scheduled order.

In order to operate the production scheduling in the productive environment, it is

important to view the scheduling of the operations of the selected orders based on the work
calendar for each production resource, following the product structure and its production
route. Therefore, the example APS system makes available the desktop Scheduling in
which it is possible to view the Gantt Graph to the scheduling generated, as shown in figure
6. In figure 6 it is also possible to view the same interface when all orders to be scheduled
are selected and the simultaneous scheduling of all selected orders is performed.

Figure 6 : Desktop Scheduling showing the scheduling of a selected order and the
scheduling of all the selected orders.
There are some schedulling rules, logical rules which determine the priority in which
the operations will be allocated to the resources, altering scheduling ans generating worse,
better scheduling depending on the situation selected.
For example, when the scheduling rule Less operation time is defined as a higher
priority, in order to schedule the production operations which may have less time, then
scheduling the operations with more time, and respecting, the precedence restrictions among
the operations. There are many scheduling rules available to be used with the example APS
system.
In APS system, it is also possible to employ mathematical optimization techniques in
order to perform the production scheduling. When this option is selected, the APS system
evaluates thousands of combinations for possible scheduling for the productive operations,
searching for a scheduling which best meets a defined optimization objective. In the example
system, the number of evaluated solutions will depend on the time made available to the
optimization and also the performance of processing of the computer in which the system is
installed. In the system used in this study, there are many possible objectives to be applied,
in order to perform the scheduling such as: minimize the delay in the delivery of all selected
orders, minimize the total scheduling time of all selected orders, minimize the inventories in
process or balance the use of the production resources.
Another feature of the example APS system is that ithas a tool which allows for the
evaluation of several scheduling parameters from the productive environment. One can
quickly analyze the performance of several liberation rules or optimization objectives, and
choose the one which best meets the specific objectives in your industry, as shown in figure
7.

FIGURE 7 Interface to evaluate the performance of an elaborated scheduling.

The scheduling method employed by the APS systems tries to meet all the
restrictions to schedule the production of a production order so that one can guarantee that
the scheduling generated is executable and can be applied in the production environment. In
the case of some personal considerations wished to added, In the example APS system, it is
possible to alter manually the generated schedule. One can alter the beginning, the end, the
duration, the resource used as well as other scheduling parameters for each production
operation scheduled, directly in the Gantt Graph bars, or by using specific interfaces to this
end.
When manual alterations are made in the scheduling generated by the APS system,
itis common for these alterations to disrespect the restrictions from the productive
environment. It is even possible to make the scheduling of the production unfeasible, mainly
when it creates conflicts on the use of resources or disregards the precedence restrictions
between the previous and following operations. Therefore, the example APS system makes
available a scheduling regeneration method which corrects the conflicts without disregarding
the alterations made.
Finally, The example APS system has several reports especially, enabling an online
view of the productive environment and allowing for the production tracking. The figure 8
shows an example of a managerial report from the detailed scheduling for the production
orders.

Figure 8 : Managerial report of the detail scheduling for a production order.

http://www.pomsmeetings.org/confpapers/011/011-0248.pdf

Strategic Planning
-Number, location and capacities of facilities,
-Investments in production and warehousing facilities
-Layout of facilities
-Distribution strategies, allocation of customers to facilities
-Outsourcing

Tactical Planning
-Procurement and production decision making
-Definition of prduction strategies
-Inventory management
-Staff and working hour management

Operational Planning
-Scheduling of the production, allocation of orders to machines
-Disposition, order processing
-Vehicle routing, vehicle loading