See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/220579402

Planning, scheduling and dispatching tasks in production control

Article  in  Cognition Technology and Work · June 2003

DOI: 10.1007/s10111-002-0117-4 · Source: DBLP

CITATIONS READS
47 19,170

2 authors:

Kenneth Mckay Vincent C.S. Wiers

University of Waterloo Eindhoven University of Technology
56 PUBLICATIONS   1,960 CITATIONS    42 PUBLICATIONS   1,073 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Human behavior in supply chain decision making View project

4C4More - Human Planning View project

All content following this page was uploaded by Vincent C.S. Wiers on 24 December 2015.

The user has requested enhancement of the downloaded file.

Planning, Scheduling and Dispatching Tasks in Production
Control

KENNETH N. MCKAY 1)
VINCENT C.S. WIERS 2)

1)
Department of Management Sciences
University of Waterloo
Waterloo, Ontario
Canada
N2L 3G1
[email protected]

2)
Institute for Business Engineering and Technology Application (BETA)
Eindhoven University of Technology
PO Box 513, Eindhoven 5600 MB
The Netherlands
[email protected]
Planning, Scheduling and Dispatching Tasks in Production
Control

Abstract:

“What is the difference between planning and scheduling?” Production control encompasses
many tasks performed by humans, three of which are: planning, scheduling, and dispatching.
In the past, the only criterion that could distinguish between the tasks was that planning is
usually on a higher level than scheduling and scheduling is on a higher level than dispatching.
Hence, the tasks are often ambiguous, unclear, and subject to speculation. There are few for-
mal studies on the actual tasks of planning, scheduling, and dispatching, and there are no
known studies that compare or discuss all three. In this paper it is argued that it is important to
understand the differences between the tasks. An action science and ethnographic case study
is presented as the empirical basis for the discussion, and the implications for decision support
system in production control tasks are presented.
Planning, Scheduling and Dispatching Tasks in Production
Control

1 Introduction
“What is the difference between planning and scheduling?” This question has been posed
many times by academics that study humans in production control and has led to long discus-
sions at conferences. One of the key issues associated with improving and advancing produc-
tion control is recognizing the different tasks that are carried out by humans. Throughout the
past century, the production control function has had some of its key components described as
planning, scheduling, and dispatching (e.g., Coburn, 1918; Knoeppel, 1920; Koepke, 1941;
Magee, 1958; Conway et al., 1967; Greene, 1970; Pinedo, 1995; Pinedo & Chao, 1999).
However, researchers that study production control tasks in practice are unable to clearly
separate a planner from a scheduler or a scheduler from a dispatcher (Crawford & Wiers,
2001).
This paper aims to give insight into the type of tasks that are carried out in production control,
focusing on planning, scheduling and dispatching. A case study that involved task analysis is
used to describe production control tasks in terms of the decision making horizon, the deci-
sion making process itself and the context. Subsequently, we translate the task descriptions to
simple information input and output characteristics. The model that we present on planners,
schedulers and dispatchers is not a cognitive task model as such; instead, the main contribu-
tion of the model lies in the recognition and delineation of the tasks using simple information
input/output characteristics. It also describes the implications on decision support systems
(DSSs) for the different tasks.
The paper is structured as follows: Section 2 elaborates on the concepts discussed in this pa-
per by giving an overview on production control concepts, functions and tasks. Section 3 con-
tains a literature review on production control task models and presents the task analysis
framework used in this paper. Section 4 contains the case study, including a description of the
action science and ethnographic methodologies used in this research. In Section 5, the simi-
larities and differences between the planning, scheduling and dispatching tasks are discussed
and presented in a model. Also, the implications on designing decision support are presented.
Lastly, in Section 6, the conclusion is presented.

2 Production control: concepts, functions and tasks

2.1 Concepts and functions
Production control concerns itself with the coordination of manufacturing activities over a
time horizon. Because the production control field is large and complex, it is often decom-
posed into several elements. One of the decompositions made is between ‘higher’ and ‘lower’
levels of production control. The higher levels usually deal with longer time horizons, and ag-
gregates the control elements. For example, a high level of production control at a truck
manufacturing company might look at trucks per month. A low level might look at engine
components to be made tomorrow and specifically what is being manufactured at two in the
afternoon. The complete coordination problem is very broad and deep.
Production control systems are often decomposed using these time and aggregation concepts
into a hierarchically organized planning and control structure to reduce the complexity of the
control problem. This approach is known as Hierarchical Production Planning (HPP). The hi-
erarchical production planning paradigm is widely used and has become an accepted planning
and control strategy for many medium to large manufacturing organizations. Bertrand et al.
(1990) describe the hierarchical production planning paradigm and they distinguish between

1
goodsflow control, which concerns planning and control decisions on the factory level, and
production unit control, which concerns planning and control decisions on the production unit
level. The goodsflow control level also coordinates the various underlying production units.
This is depicted in Figure 1.

goodsflow control

production unit control production unit control

decoupling decoupling decoupling
point A point A point
D C D C

Figure 1: Goodsflow control and production unit control (adapted from Bertrand et al., 1990)

A production unit is an outlined part of the production process of a company. A production

unit produces a specific set of products from a specific set of materials or components, with a
specific set of capacity resources. Dependent on the complexity of a production system, a
production unit can be a single machine, a group of machines with the associated personnel,
or an entire factory. Production control decisions are often decomposed to conform to the de-
composition of the production system.
Production control at the goodsflow level is associated with planning, whereas production
control at the production unit control level is associated with scheduling. Planning and sched-
uling are also well-known concepts in the operations research literature. In the operations re-
search domain, planning can be defined as: applying ordering constraints on actions (e.g.,
Burgess & Steel, 1996). Scheduling is defined as: allocating jobs to resources in time (Con-
way et al., 1967).

2.2 Functions and tasks

The concepts of planning, scheduling and dispatching in production control are defined more
or less as black boxes. Academic and research oriented literature, such as Bertrand et al.
(1990), define material planning in terms of the input, i.e. the master production schedule, the
bill-of-materials and inventory data; output (e.g., timed-phased requirements records); and ob-
jectives (e.g., desire to provide the right part at the right time). However, apart from the basic
logic, there are no descriptions on how this function should actually be carried out in practice,
i.e. how the black box performs its task.
In the industrial sector, the widely used Manufacturing Resource Planning (MRP II) concept
identifies the following functions: master production scheduling, rough-cut capacity planning,
material requirements planning, detailed capacity planning, work order release, and shop floor
control (Plossl and Wight 1967, Orlicky 1975, Vollmann et al., 1988). Apart from the basic
logic, the literature on MRP II does not describe how each function should actually be carried
out in practice.
In many production settings, this lack of clarity and prescription leads to the situation where
almost every firm or plant interprets the meaning of planning, scheduling, and dispatching
differently – task assignments, naming, organizational structure, and so forth. This inconsis-
tent and imprecise structure within and between firms means that the production control tasks
are difficult to study scientifically: observe, describe, or compare. The task combinations ap-
pear almost limitless. For example, from a production control point of view, a machine opera-
tor may take on the role of dispatcher, or one person may be planner and scheduler and dis-

2
patcher. The employees may also expedite shipping, contact suppliers regarding material arri-
val, audit inventory levels, and reconcile shop floor results to name just a few of the many ac-
tivities planners and schedulers have been observed to do. Figure 2 illustrates some of the
possible variations in task content.

Task 1 planning

scheduling
warehouse material
management
inventory management
control
shopfloor Task 2
execution

Figure 2: Possible variations in production control tasks

The definitions of planning, scheduling and dispatching from the production control field do
not clearly indicate how the functions should be carried out. For example, if one human does
both planning and scheduling, how to clearly recognize the planning and scheduling part of
the job is not defined? What is the difference between either ‘applying ordering constraints on
actions’ and ‘allocating jobs to resources at a specific point of time in the future’? In other
words, what is the difference between planning and scheduling? And where does dispatching
fit in?

3 Production control task models

3.1 Literature review
There have been several field studies on humans that perform production control functions (an
overview is given in Crawford & Wiers, 2001). However, researchers that study production
control tasks in practice rarely separate a planner from a scheduler or a scheduler from a dis-
patcher. For example, Sanderson & Moray (1990) use scheduling models to understand how
time pressure affects tasks. Nevertheless, in the study, humans are compared with dispatching
rules. Sanderson (1989) reviews the human planning and scheduling role in advanced manu-
facturing systems, and indicates that the focus for her review is scheduling and dispatching.
Sanderson defines scheduling and dispatching as “relatively short-term activities determining
how and when a product will be produced.”
There are few studies on production control tasks that aim to construct a comprehensive task
model, and the studies fail to build upon each other (Crawford & Wiers, 2001). The majority
of studies appear to start from a new perspective, disregarding previous work and have failed
to create a firm foundation for a definitive model for production control tasks.
In some of the early studies, the focus was to construct a mathematical model to describe the
scheduler’s decision behavior (e.g. Dutton, 1962, 1964; Fox & Kriebel, 1967; Hurst and
McNamara, 1967; and Dutton & Starbuck, 1971). Whereas these authors all claim to study
scheduling problems, the studies contain the following variety of tasks: scheduling corrugated
box fabrication, production sequencing in a shoe-box plant, planning a woollen mill, produc-
tion sequencing and machine time prediction in a textile plant. Thurley and Hamblin (1962)
conducted five small case studies in which a variety of manufacturing situations were ana-
lyzed. Their focus of research was the fact that technological, human and organizational fac-
tors all influenced the behavior of production supervisors who also carried out planning and
scheduling tasks as part of their roles. However, they did not make a clear distinction between
the planning and scheduling tasks. Other knowledge elicitation studies have been carried out

3
under simulated or controlled conditions (Tabe et al, 1988; Lopez et al, 1998) or as limited
field studies (Duchessi & O’ Keefe, 1990; Norman & Naveed, 1990). The findings tend to be
specific to particular research conditions and therefore cannot be applied generically.
Focusing on the human and system interaction aspects, McKay (1992) attempts to explain
how the human is crucial to production control decision making and proposes a framework
that captures the type of man-machine control that is necessary. Subsequent research by
McKay et al. (1995a, 1995b) addressed other issues related to the human factors of schedul-
ing, many of which deserve further research. One of the few attempts to build a task model for
a production control task is described in Sanderson (1991). She constructs a framework for
the Model Human Scheduler (MHS) that consists of twenty–seven production rules linking
different types of scheduling activities. However, this line of research has not been pursued.
In studying tasks in production control, Crawford (2000) distinguishes between three roles: an
interpersonal role, an informational role and a decisional role. Furthermore, she describes
three tasks for a scheduler: formal tasks, housekeeping tasks and compensation tasks. How-
ever, in this study the distinction is not clear between the three levels of decision making. Fur-
thermore, it is not clear if the model needs to be adapted if an individual finds themselves in a
multi-tasked situation, e.g. formal and housekeeping.

3.2 Framework for task analysis

From a task characteristics point of view, there seem to be similarities between planning and
scheduling on the one hand, and scheduling and dispatching on the other. These similarities
lead to the fact that models developed by earlier studies are not strictly confined by the au-
thors to either planning, scheduling or dispatching tasks. The only distinguishing characteris-
tic for planning, scheduling and dispatching that can be derived from existing literature is that
planning is done on a higher aggregation level and dispatching on a lower aggregation level,
with scheduling in-between.
Because this paper attempts to describe the difference between planning, scheduling and dis-
patching, the focus will be on the interrelationships between the tasks. More specifically, this
paper focuses on the aggregation level of the input and output of the tasks to distinguish be-
tween planning, scheduling and dispatching. Three task characteristics will be used as a
framework to analyze the tasks:
• Horizon and Timing. The decision making horizon can be measured in years, months,
weeks, days, hours, and minutes. The timing of decisions can be continuous or periodic.
• Decision Making. The type of decisions that are made in the task can be defined, and the
amount of autonomy that is typically present in the task can be specified.
• Context. The decision making context can define the when, what, and why that helps to de-
fine the functions and tasks that a planner, scheduler or dispatcher has to interact with.

A possible fourth task characteristic would be performance; however, this characteristic is

found to be closely related to the decision making in the production control tasks. Therefore,
descriptions about task performance are embedded in the decision making category.

4 Case study
4.1 Background
A long term field study on production control has been conducted in one plant. Over a six
year time period, the study has focused upon dispatching, scheduling, and planning within the
plant. The horizontal scope within the plant ranged from the receiving of incoming raw mate-
rials to the shipping of finished goods. The study also occurred at different vertical levels
within the plant: from the shop floor to management policies.

4
The following subsection overviews the methodology and the case study site. The focus of the
following subsections is not on the methodology per se or the various detailed observations
that accumulated over the study period; instead, the purpose is to establish the general field
methods that allowed the aggregate view in Section 5 to be prepared. For those interested, de-
tailed observations and aspects on production control have been reported elsewhere (McKay,
1992; McKay et al., 1995a; Wiers & van der Schaaf, 1997; McKay & Wiers, 2001; Wiers,
2001).

4.2 Research methodology

4.2.1 Introduction
The research methodology was a combination of action science and ethnography. As the re-
search activity was extended and broad in its scope, the two methods were used at different
times for different goals. In both, the researcher was involved in the situation and was a full
participant with a role to play.
In the action science vein, the researcher asks “How does it happen to be?” and “How might
we transform what we discover?” (Argyris et al., 1985, p. 228). The action science research
may also have a priori theories or issues that are being used and tested. In contrast, the focus
of the ethnographic approach is to learn from the subjects, avoiding judgement and biases,
and to relatively passive in the situation (Spradley 1979).
Case studies of the type described by Yin (1989) are typically of a limited duration with a sin-
gle snapshot, or with a restricted number of visits. Action science studies are normally more
extensive, but rarely extend to the duration typical of ethnographic studies, which try to cap-
ture full life cycles and the related dynamics.
Each technique, action science or ethnographic analysis, is suitable for different research ob-
jectives and situations. At the study site, the activities were wide-ranging horizontally and
vertically throughout the plant ranging from shipping to receiving and from senior manage-
ment strategies to shop floor operations. As a result, a mixture of action science and more tra-
ditional ethnographic research was used. Action science was used in the study of daily opera-
tional decisions in one area of the plant with a study and transform role. Ethnographic analy-
sis techniques were used for researching planning, scheduling, and dispatching throughout the
plant with an integration role.

4.2.2 Action science

An action science oriented study can be considered a fourth type of case study (Argyris et al.,
1985), where limited case studies are usually either exploratory, descriptive, or explanatory
(Yin, 1989, p. 13). Gummesson (1988) discusses Yin’s categories and this fourth style, and he
claims “Action research is the most demanding and far-reaching method of doing case study
research.” (p.99). Subsequently, Gummesson lists eight characteristics exhibited by action
science research. In action science, the researcher may do some or all of the activities related
to basic case study data collection and analysis, but also be part of the actual process, partici-
pating in the case, and partially responsible for any success or failure (e.g., Jorgensen 1989,
Easterby-Smith et al., 1991).
One of the goals of action science is to discover what practical information is used to trigger
and control the actions. This is done by participating in the process. Subsequently, this infor-
mation is then used to transform the situation using a theoretical foundation. The action sci-
ence methodology can be used to discover new understanding of the triggers and issues in a
situation. The larger phenomenon might be previously identified, but the details not under-
stood. Action science can also be used to test existing theories of cause and effect. In the ac-
tion science methodology, an existing situation is identified, prescriptions are in place, and the
process of study and transform scientifically applied, recorded, and analyzed. If the research

5
effort is merely applying widely existing knowledge and techniques to the situation without a
scientific agenda, the activity may resemble consultancy with little academic value.
At the study site, the action science approach was used to study and then transform the short
term scheduling activities for part of the plant. Using concepts and theories of task allocation
and context-sensitive software engineering, a single integrated task was decomposed into two
tasks with an added worker, granularity of decision elements changed, decision horizons al-
tered, work visibility tuned to remove irrelevant items, and specialized decision support sys-
tems created for the two workers. This transformation activity, which took approximately two
years, provided the opportunity to start analyzing and dissecting what was truly different be-
tween dispatching and scheduling.

4.2.3 Ethnographic approaches

Ethnographic approaches are intended to be less invasive than action science activities. With
the intent to learn from the subjects, there are three phases of ethnographic research com-
monly undertaken: observation (non-participant to full participant involvement), interviews,
and quantitative analysis based on questionnaires, structured interviews or data collection (see
Spradley, 1979; Grills, 1998; and Schensul, 1999). When a colony or group is studied, the
methodology progresses from the observation level to the quantitative analysis.
Following the action science research where tasks were radically transformed, we were given
the opportunity to replace an existing set of decision support tools with an integrated tool. In
this case, the goal was not to radically transform the task, and a different research approach
was needed. The replacement of the tools was completed over a two year period, during
which the first decision support system continued to be used and supported. Both areas have
continued to use the various decision support tools and have received ongoing support in the
two years since.
Both activities, the transformation and integration roles, provided an opportunity to observe
the scheduling tasks for an extended time period (six years) and to see the longitudinal as-
pects. It is our view that the scheduling task exists within a culture, and is itself a culture with
behaviors, terminology, norms, expectations, interactions and interrelationships. This is simi-
lar to the view taken by Hutchins (1995) in studying the cognition elements of navigators on
naval vessels. However, the navigation task is at one end of the cognitive planning spectrum,
while factory dispatching and scheduling is at the other: approaching dozens and possibly
hundreds of harbors each day in a continuous fashion does not typically happen for large mili-
tary vessels, as does the job scheduling activity in a factory. Neither does it have the con-
straint elasticity that factory planning and scheduling has: the wharf cannot be moved quickly
or another channel dug in real time. Furthermore, the navigators have one main objective,
docking the vessel, with many factors to control, while a factory scheduler has many inde-
pendent objectives. Nevertheless, there are similarities in that cultures exist that define the
task, terminology is heavily used, and there are many expectations in the workplace that guide
behavior and interpretations. In both studies, production control and naval navigation, ethno-
graphic and field study methods were used to observe and understand what the task really
was, what the people really did, and why they did it.

4.3 The factory

4.3.1 Primary process

A brief description of the field study environment is given below; a more extensive descrip-
tion is given in McKay & Buzacott (2000). The site, which we will call ACME, is in the mid-
dle of the automotive supply chain, has approximately one dozen final assembly lines and
ships roughly four dozen final items in a just-in-time fashion to several different customers.
ACME runs a three shift operation and often plans production for the weekend. The flow time

6
is measured in minutes and production is in hundreds per shift for each line. The factory is
composed of two major areas: a flow shop and a job shop.
Production is pulled through the plant from the finished goods inventory. The assembly area
builds on dedicated resources and fills a target level of on-hand finished goods. The job shop
works in a batch mode with a cyclic pattern. The area is made up of cells and each cell has
one or more machines. The capacity in the job shop is theoretically balanced to that of the
flow shop.
There are a number of characteristics about ACME’s job shop which are not found in tradi-
tional job shop problem formulations (e.g., Conway et al., 1967; Baker, 1974; Pinedo, 1995;
Pinedo & Chao 1999):
• Virtual flow lines
• Crewing issues
• Independent setup phase
• Transfer within batch capability
• Ability to pause a job indefinitely and move resources
• Independent but co-dependent parts

The above issues complicate the decision process beyond that of simple job characteristics be-
ing matched with machine availability, which is the traditional view of job scheduling. With
the exception of the job shop, the plant’s production system is relatively straightforward.

4.3.2 Production control

Production control in the plant focuses on the following parameters:
• Setups (e.g., unnecessary setups caused by preemptions, miscues in communication)
• Direct manpower levels (e.g., overtime)
• Indirect manpower associated with production (e.g., support groups)
• Inventory (e.g., too much, too little)
• Shipping (e.g., expediting deliveries at a higher cost)

Missing from the above description is any mention of ‘traditional’ production metrics such as
minimizing lateness, weighted tardiness, the number of late jobs, and so forth. In lean manu-
facturing as practiced in the automotive sector, these metrics are not the major ones to track or
focus upon. For example, in the factory we are describing, the concept of ‘lateness’ is not
used in planning, scheduling or dispatching. Lateness is not part of the final schedule. If there
is any lateness, it is not planned and is considered totally unacceptable with significant penal-
ties and costs. At a tactical level, metrics such as flow time are tracked by management, but
are not explicitly addressed by the production control process.

4.3.3 Roles and Tasks

At ACME it was fairly easy to identify planners, schedulers, or dispatchers, because a job ti-
tle of planner existed in the plant and was held by four individuals. In the job shop, there was
a planner, scheduler and dispatcher, although sometimes they did each other’s job, sometimes
rotated regularly through the positions on shift work, and sometimes extended their decision
making into another area. They also substituted for each other when vacation or illness dic-
tated. On a daily basis, one was charged with dispatching, one with scheduling, and one with
planning. Contrarily, in the flow shop, there was one planner who always did a combination
of planning, scheduling and dispatching. The tasks will be described in detail in the next sec-
tion.

7
4.4 Task analysis

4.4.1 Planning

Horizon and Timing

Since the early 1900s, the levels of granularity for planning have been influenced by a number
of factors, such as: workers prefer a weekly shift pattern, management requires aggregation,
fiscal reporting must be done in months, quarters, and years. These forces have often led to a
so-called bucket brigade, which means that demand and supply are viewed grouped by weeks
and months. Like other planners we have studied, the planners at the plant deal with horizons
that are measured in years, months and weeks. However, monitoring activities that relate to
the forecast and incoming orders are carried out at a day or week level. The monitoring is im-
portant with today’s Electronic Data Interchange (EDI) and pull situations: sudden shifts in
demand can wreak havoc with carefully orchestrated resource allocation. However, it can also
provide clues about the future.
The timing of planning decisions depends on the production control situation. In a make-to-
stock situation, the planner needs to make a decision for each time period for each product. In
a make-to-order situation, the planner needs to make a decision about any changed and new
work. A daily decision making activity of the planner lies in updating the plan as a result of
changes in capacity or demand. The precise frequency of decision making depends on when
the customer orders appear in the system that the planner uses, and how often the system is
updated with production data. It is common with firms in the automotive sector to have a mix
of each type of situation. For example, prototype and service parts are more of a make-to-
order situation whereas the normal pull from an assembly plant has a make-to-stock profile
with multiple year forecasts provided by the customer.

Decision Making
The planner for each area attempts to balance demand with available capacity using inventory
as a buffer. Inventory is used to stabilize production and the desired levels are allowed to float
between a minimum and maximum. Major capacity issues are identified that require substan-
tial amounts of overtime, shift changes, or subcontracting. In cases of imbalance, the planner
will try to level or change demand or supply.
Planners usually have more autonomy regarding supply and demand than the scheduler or
dispatcher. The planner is in many cases able to change the demand volume, requested prod-
uct, due dates, and shipments. The total volume may not be altered, but the mix and delivery
schedule can be adjusted for what is on hand at the customer, what is in the pipeline to the
customer, and what is currently on hand in finished goods inventory at the factory. It is possi-
ble to intelligently manage the inventory at each level and use one level to another level’s ad-
vantage in a temporary fashion. That is, targets are set for inventory levels but they do not
have to be followed blindly in a mechanical way.
The supply side can also be manipulated by the planner. The planner might negotiate addi-
tional manpower, different shifts, scheduled overtime, or resource changeovers. In previous
decades when significant levels of inventory existed throughout the supply chain, the fre-
quency and type of manipulations were less. However, the lean targets for manufacturing
force the planners to be more proactive and on top of the situation daily.
When the complexity of the planning domain (i.e., number of parts, periods) increases beyond
the planner’s cognitive abilities, three techniques are commonly used. First, information is
aggregated into larger buckets: e.g., planning for the week instead of by the day. Second, ad-
ditional planners are used and the work is distributed. Third, the planner can choose to ignore
the majority of the daily task and focus on critical spots.

8
Context
The planners have regular interaction with managers to plan the work force, set production
control policies, acquire capacity, and the like. Decision making about the future is one part of
the task: the planner is also responsible for tracking and analyzing production as it compares
to the plan. Various reports are fed back to management daily, weekly, and monthly. There is
a regular monthly planning cycle for firming up the next period and to account for any
changes in the twenty-week horizon. Periodically through the year there are major planning
activities for yearly, three year, and five year plans. For the major planning events, the planner
is expected to generate multiple scenarios quickly and accurately.
Changes made by the planner are often not felt by the scheduler or dispatcher. The planner in-
teracts with the scheduler and dispatcher when changes in demand or capacity cannot be dealt
with in the mid or far future.

4.4.2 Scheduling

Horizon and Timing

The horizon of the scheduler is determined by the amount of uncertainty, but usually one
week. If the amount of uncertainty increases, the horizon used by the scheduler decreases, and
vice versa.
The scheduling task can roughly be split into two parts: constructing the schedule and reactive
scheduling. Constructing the schedule is usually done once or twice per week if the horizon is
one week. This is done by moving work around, making sequences, and smoothing loads as
the customer requirements are loaded in and become visible. The rest of the week is spent re-
acting to requested maintenance downtime, major repair plans, and any conscious build ahead
or push out of demand. Decision making is almost non-stop in terms of moving, swapping, in-
specting job attributes, and the like.

Decision Making
The scheduler tracks any resource and material issue, change in capability or capacity that can
result in significant constraint changes. Ideally, the scheduler does not plan for head-to-tail re-
lationships on all work and allows certain safety time or buffers wherever there is a potential
risk situation; this is the mechanism that allows some simplifying assumptions to work. For
this reason, the scheduler orchestrates resource allocation to accommodate special require-
ments, such as preventive maintenance, service parts, emergency builds, vacation season.
A scheduler does not simply start at the beginning of the problem space. He or she will priori-
tize issues according to the types of constraints involved (Wiers, 1997):
• work that is critically constrained, i.e., it endangers the goals of the scheduler;
• work that is tightly constrained, i.e., it has little alternatives;
• work that is extensively constrained, i.e., it needs to obey to many (interrelated) con-
straints; and
• work that is stochastically constrained, i.e., history has proved that a specific set of work
is often troubled by disturbances.
To create a feasible load, the scheduler will focus on manipulating the demand instead of the
available capacity. This is done by working with the planner and purchasers, or for example
by batch slitting. If the demand pattern cannot be tweaked, the scheduler will try to use alter-
native resources, overtime, or negotiate relatively easy changes to the resources.
It is hard for the scheduler to see the quality of the scheduling decisions because of the time
delay and the number of changes that occurs. Moreover, it is hard to assess what factors have
influenced a certain performance level to what extent (see also Gary et al., 1995; Stoop,
1996).

9
Context
The scheduler is responsible for firming up plans and creating a feasible sequence for the dis-
patcher. The dispatcher is looking to the scheduler for work to consider, and the planner is
concerned about impacts on the plan caused by any scheduling changes. Schedulers get pres-
sure from the dispatcher and the planner. Schedulers rarely get much pressure from the opera-
tions people, except in cases where dispatching is being carried out by operations people.
Management may also be involved in scheduling, but the intensity is less than at either the
planning or dispatch level. In some ways, scheduling is almost an invisible layer or activity.
The dispatcher and scheduler interaction can be tense if there is not enough work or the right
work that is ready to be dispatched. For example, the dispatcher may be dealing with a bottle-
neck and the scheduler has not moved in or firmed up the work necessary to keep the bottle-
neck busy. There can also be tension with support groups if the scheduler moves work in
ahead of material or tooling expectations. During periods of less-than-peak capacity utiliza-
tion, changes made by the scheduler should not affect the planner to any great degree. How-
ever, when the shop is running behind and everything is ‘tight’ on the schedule, any change
made by the scheduler will affect the planning.

4.4.3 Dispatching

Horizon and Timing

The dispatcher’s horizon is measured in minutes, hours, and days. The dispatcher usually has
two jobs planned in detail for each resource: the current one and the next one. Many of the
jobs will run for two to three days, and decisions beyond this are considered as scheduling in
the factory.
Dispatching decisions are made continuously each day. If the dispatcher is the machine opera-
tor, the decisions are made when the current job finishes and the next must be started. When
the dispatching is centralized and the dispatcher has multiple resources to contend with, then
multiple decisions are made throughout the day. In a medium job shop with about sixty re-
sources, a dispatch decision can be made every five to ten minutes throughout the day.
Another trigger for a decision by the dispatcher is when the workforce arrives on the scene. In
this case the dispatcher might decide what jobs to start with. Moreover, if there are more ma-
chines than people in a specific shift, the dispatcher also decides what to work on.

Decision Making
The usual sequencing decisions that are made by the dispatcher include setups, resource
availability, job duration, and due dates. The dispatcher must deal with preemption, running
short, running over, machine breakdowns, personnel problems or absenteeism, current state of
the tooling, and current availability of material. In making decisions, there are various costs
and issues that must be considered: cost of mobilization, increased risk of quality problems,
extended processing times, possible damage to the process, destabilization of existing system
and processes, and the cost of reverting. Moreover, due to the time pressure to make a deci-
sion, the options will not be explored in-depth.
The dispatcher must deal with very detailed data, including any information that can affect the
choice of resource, the duration of the operation, and the quality of the result. The range of
data used by the dispatcher is exceptionally broad covering environmental factors (e.g.,
weather), cultural phenomena (e.g., impact of the holiday season on absenteeism), worker mo-
rale (e.g., who is likely to give that extra effort or work overtime), recent performance (e.g.,
who did what the last time and how well it was done), and the normal engineering data (e.g.,
standards, material specifications, processing descriptions). Any piece of information, no mat-
ter how small or seemingly innocent, may have relevance. Moreover, the dispatcher is con-
stantly gathering gossip, tidbits of conversations, and is generally being nosey.

10
The dispatcher has limited influence on the demand and some on the available capacity. In the
short term, the dispatcher can do little to a demand requirement in a one-of-a-kind situation
since there are no buffers or safety-stock to use. In other situations, the dispatcher may be able
to contact the customer and delay or advance work. Furthermore, the dispatcher may be able
to exploit elasticity in production quantities and safety-stock. In this case, there will be an im-
pact in the future, but the day will be saved.
Usually, the dispatcher has more degrees of freedom in the supply side, which are usually ex-
plored before the demand side. There might be workarounds to make a product, for example:
using a previous generation machine or adding manual operations. Other manipulations in-
clude: splitting or merging operations, adding or altering equipment, using substitute material,
making substitute material from other material, extending shifts for extra capacity, subcon-
tracting operations.

Context
The dispatcher has the most immediate context, because the decision making of the dispatcher
immediately affects the use of personnel and resources. The dispatcher works with the lead
hands, supervisors, foremen and operators to review options, analyze the impact on other
work scheduled, and decide on the next steps. Any loss on a bottleneck resource will delay the
whole process and this responsibility falls to the dispatcher. The dispatcher must also consider
the impact decisions on the scheduler and planner, especially in a tightly packed plant where
decisions spread like a ricochet.

5 Planning, Scheduling, Dispatching

5.1 Classification of Differences
Planners, schedulers and dispatchers represent three levels of production control. This article
has demonstrated that there are significant differences between the planning, scheduling and
dispatching task. Table 1 summarizes the differences found in the task analysis of Section 4.4.

Availability of Interaction with

Horizon Time pressure
information lower level
Planner Multiple, bucketed High Low Low

Scheduler Single, real-time Lower High Higher

Dispatcher Single, real-time Low n.a. High

Table 1: Differences between planners, schedulers and dispatchers

First, while schedulers and dispatchers are faced by similar demand and supply decisions
within their time horizons, the planner simultaneously tracks daily issues, prepares weekly
summaries, monthly plans, yearly plans, and so forth. The decisions of the planner are formu-
lated as buckets of work in a specific time period, whereas schedulers’ and dispatchers’ deci-
sions are formulated as specific tasks in real-time.
Second, the planner has much more information regarding demand than the lower production
control levels: for example, when the customer or sales has put extra padding in or is building
ahead. The planners also know when the customer is planning down time, line changes, and
special try outs. The majority of this additional information comes from personal contacts that
are not available to the scheduler of dispatcher because of their different contexts.

11
Third, in Section 4.4.1 it was stated that changes made by the planner are often not felt by the
scheduler or dispatcher. Because the planner works with buckets, the relationship between the
planner and the scheduler are defined clearly and the planner can work in relative isolation
from the scheduler. However, the scheduler is more actively concerned with the dispatcher: he
or she tracks any resource and material issue, change in capability or capacity that can result
in significant constraint changes that are not easily handled by dispatching. The scheduler
works with days and weeks, and assumes that ‘minor’ issues such as tool repairs can be ig-
nored and left to the dispatcher.
Fourth, there are large differences in the pressure that is related with the tasks. For the sched-
uler, the future is the focus, and hence the immediate pressure is not as great as for the dis-
patcher. The dispatcher must function like a fire department with everything ready for deci-
sion making continuously. The dispatcher gets faster and more vocal feedback on decisions
that planners and schedulers and this changes the operational context. For example, the dis-
patcher must explain any ‘poor’ decisions that affected the afternoon or evening shift, explain
why they were unable to see the difficulty in advance, and explain why they are changing the
plans for today they made yesterday. The planner has somewhat less pressure because there is
much more time to solve problems, and problems can be more easily escalated to the respon-
sible managers.

5.2 Task delineation model

Based on the input and output characteristics of the task, it is possible to draw up a task de-
lineation model that answers the question: “What is the difference between planning, schedul-
ing and dispatching?” In other words, we are not delivering a comprehensive cognitive task
model as a result of the research described in this paper; rather we present a simple model that
researchers and practitioners in the production control task field can use to distinguish be-
tween the planning, scheduling, and dispatching task.
As stated in Section 3, it can be difficult to identify the planning, scheduling or dispatching
components in what the human is actually doing. Existing literature teaches us that planning
is ‘more’ aggregated than scheduling, and scheduling is ‘on a higher level’ than dispatching.
However, these task characteristics are of a qualitative nature and cannot be measured objec-
tively. This is not a problem if the tasks exactly follow the production control structure in a
specific company, and if the production control structure in practice is in strict accordance
with theory. However, most situations do not seem to follow this stereotype, and even theory
itself is inconsistent in using the terms ‘planning’ and ‘scheduling’. Hence, for most situa-
tions, there are no models that indicate at which point precisely the planning task ends and the
scheduling task begins. Therefore, any interpretation of a task by using these characteristics is
arbitrary.
To distinguish between the planning, scheduling and dispatching tasks, task characteristics are
needed that can clearly be applied and interpreted. Based on the task analysis findings pre-
sented in the previous subsection, it is possible to extract two attributes that help decompose
the decision processes into well-defined categories. The two attributes are the input drivers
and output representations of the task. For example, the input of the dispatcher is based on
today’s facts, instead of forecasts as in the planner’s and scheduler’s case. Also, the output of
the planner is represented by volumes in buckets, whereas the scheduler and the dispatcher
work with jobs in real-time. A task delineation model based on these two task attributes is
shown in Table 2.

12
 Input driven by: Output represented by:

Planner Volumes / bucket

Expectations of
Scheduler the future
Orders, assignments, or
jobs / real-time
Dispatcher Reality
Table 2: Delineation of planning, scheduling and dispatching task

As shown in Table 2, there is overlap between planning and scheduling on the one hand, and
scheduling and dispatching on the other. However, when the two task characteristics are com-
bined, it is possible to distinguish between the three tasks unambiguously.

5.3 Implications for decision support

The results discussed in sections 5.1 and 5.2 assist with understanding what an individual
does and what is expected of them. This is useful for task and role analysis and for categoriz-
ing what a scheduler or dispatcher does daily, weekly, monthly, and so forth. Once the tasks
are more clearly separated, it is possible to consider the how decision support systems can be
designed to better support the various tasks and roles; especially when individuals have multi-
tasked positions.
It is obvious that the differences between planning, scheduling and dispatching result in dif-
ferent requirements for decision support. When the task is clarified using the model in section
5.2, the types of interactions associated with each task are isolated and can then be further
analyzed. To organize the requirements and to understand them better, the requirements can
be grouped according to a decision taxonomy such as introduced by Wiers (1997). In this tax-
onomy, there are four key task characteristics that set the requirements for decision support
systems in production control tasks:
• The amount of autonomy that should be supported by a decision support tool for a specific
task depends on the uncertainty and flexibility of the lower production control task levels.
If the lower control levels are stable and predictable, the higher control level can get more
autonomy.
• The transparency that a decision support system should offer results from the extent to
which the task is critical and ill-defined. When the task is more critical and ill-defined, the
need of the human to be in control and thus the required transparency of the tool increases.
• The level of support that a decision support system can offer depends on the amount of the
real-world that can actually be modeled in the system. If there are many occurrences where
the human must react manually, the level of support of a system should be lower.
• The information presentation by a decision support system should match the human’s
representation of the problem domain. Furthermore, aggregation should be used by the sys-
tem to offer an overview in complex production control domains.

The four characteristics of a decision support system for production control tasks are trans-
lated to the tasks that are analyzed in the case study in the table below.

Planner Scheduler Dispatcher

13
 Planner Scheduler Dispatcher
Schedulers have much
The dispatcher’s deci-
As planners have rela-
interaction with dis-
sion is followed exactly
tively little interaction
Autonomy patching, and their
by the shopfloor, there-
with scheduling, their
autonomy is therefore
fore their autonomy is
autonomy can be high.
limited. high.
The planning task is The dispatching task is
The scheduling task is
less critical than the more critical than the
more critical than the
Transparency other tasks, so the scheduling task, so
planning task, so trans-
transparency of the transparency must be
parency must be higher.
DSS can be lower. higher.
The level of support The level of support has
The level of support has
can be relatively high, to be intermediate, be-
Level of to be low, because
because the domain is cause not all informa-
support much information is not
well-specified (infor- tion is formally avail-
readily available.
mation is available). able.
The main presentation
of the DSS should a list
of jobs released by the
The main presentation
The main presentation scheduler to choose
Information of the DSS should show
of the DSS should be a from. An overview is
presentation multiple buckets in a
Gantt-chart. usually not needed be-
table.
cause the set of jobs
considered is relatively
low.
Table 3: Requirements for decision support for the planning, scheduling and dispatching task

Research in production control often implicitly assumes that the functionality of a decision
support tool will follow the requirements as specified in models, such as described in the table
above. This assumption usually holds in cases where software is custom built for a specific
situation with unlimited funds (Wiers, 2002). However, in practice, companies implement
commercially available standard software packages. Decision support for production control
tasks is typically offered by Enterprise Resource Planning (ERP) systems (Markus & Tanis,
2000), which are based on the MRP II paradigm, and Advanced Planning and Scheduling
(APS) systems (Stadtler & Kilger, 2000). These standard packages are configured to the char-
acteristics of the company, and the functional flexibility of these systems is limited (Wort-
mann, 1992).
There are countless possible applications of ERP, APS and other systems to support produc-
tion control tasks. Therefore, this paper presents one typical configuration of an ERP and APS
system and will discuss possible pitfalls that exist in implementing these systems in light of
the research results. The figure below shows a typical configuration of an ERP and an APS
system.

14
 MRP Master Planning
APS
Material
MES Production detail
Requirements
scheduling
Planning

Dispatching
Figure 3: Typical Configuration of ERP and APS

Planning tasks in industrial companies are often supported by ERP systems, which for pro-
duction control are based on the MRP II concept. APS systems usually have a module for
production scheduling. The dispatching is often done on the shop floor using a so-called
Manufacturing Execution System (MES). This typical decision support structure for planning,
scheduling and dispatching has the following weaknesses:
• Difference between planning and scheduling is unclear. The output from planning is often
fed into the MRP algorithm which performs a time-phased material explosion based on the
assumption of infinite capacity. The outcome is a draft schedule that is proposed to the
scheduler. However, the parameters used by MRP are often not correct (see for example
den Boer, 1994). Consequently, the scheduler changes the MRP output and is thereby con-
ducting planning task activities. This weakness is illustrated by the fact that the question
“What is the difference between planning and scheduling?” is very hard to answer in prac-
tical situations.
• Dispatching is underestimated. Both in ERP systems that are based on the MRP II frame-
work as in APS systems, there are no specific modules for dispatching. Apparently, either
it is assumed that the scheduler also does the dispatching, or the dispatching is done by op-
erators on the shop floor.
• Scheduling systems are not transparent and their level of support is too high. Most sched-
uling systems assume that the scheduler has a large autonomy. Therefore, these scheduling
systems contain algorithms to construct a ‘optimal’ schedule. This leads to a schedule that
is far too prescriptive for the dispatcher and by automatically generating a schedule the
system becomes opaque to the human.

The above mentioned weaknesses of current state-of-the-art decision support systems lead to
the observation that the differences between the planning, scheduling and dispatching tasks
are under estimated and that the link between the tasks is unclear. The task delineation model
that has been presented in this paper aims to give researchers and practitioners that deal with
production control task a clearer definition of planning, scheduling and dispatching tasks. A
detailed description of an implementation of a decision support system that is based on the
concepts presented in this paper is described in McKay & Wiers (2003).

6 Conclusion
In this paper it has been stated that a clearer separation can be made between the planning,
scheduling and dispatching tasks in production control. While job titles and organization
charts may not be clear and individuals may do all three, it is possible to use time horizons,
types of decisions, and the context to make the distinction.
The action science and ethnographic case study described in this paper provided the mecha-
nism for observing and participating in the situation. Software tools for planning, scheduling,
and dispatching were created, some tasks were re-designed, and some production control

15
structures were overhauled completely. It was from these observer and participant activities
that the taxonomy proposed in this paper arose.
The classification of tasks aims to deepen our understanding of production control, to help re-
searchers in the human factors in production control, and to facilitate improvements of pro-
duction control performance by supporting humans with adequate decision support tools.

Acknowledgements
This research has been supported in part by NSERC grant OGP0121274 on Adaptive Produc-
tion Control. The senior plant management and the individuals in the production control de-
partment are gratefully thanked for the support shown during the project.

List of Acronyms
APS Advanced Planning & Scheduling
DSS Decision Support System
EDI Electronic Data Interchange
HPP Hierarchical Production Planning
MES Manufacturing Execution System
MRP Material Requirement Planning
MRP II Manufacturing Resource Planning
WIP Work-in-Process

References
Argyris, C., Putnam, R., & McLain Smith, D. (1985). Action Science. San Francisco: Jossey-
Bass Publishers,.
Baker, K.R. (1974). Introduction to Scheduling and Sequencing. New York: Wiley.
Bertrand, J.W.M., Wortmann, J.C., and Wijngaard, J.(1990). Production control: A structured
and design oriented approach. New York: Elsevier Science Publishers.
Boer, A. A. A. den (1994). Decision Support for Material Planners. Production Planning &
Control, 5(3), 253–257.
Burgess, A., and Steel, S. (1996). Quantification in Causal Link Planning. In Proceedings of
the 15th Workshop of the UK Planning & Scheduling Special Interest Group, UK,
November 21–22 (pp. 57–72). Liverpool, UK: Liverpool John Moores University.
Coburn, F.G. (1918). Scheduling: The Coordination of Effort. In I. Mayer (Ed.), Organizing
for Production and Other Papers on Management 1912-1924 (pp. 149–172), Easton:
Hive Publishing.
Conway, R.N., Maxwell, W.L. & Miller, L.W.(1967). Theory of scheduling. Reading, MA:
Addison–Wesley.
Crawford, S. (2000). A Field Study of Schedulers in Industry: Understanding their Work,
Practices and Performance. PhD thesis, University of Nottingham, UK.
Crawford, S., and Wiers, V.C.S. (2001). From anecdotes to theory: reviewing the knowledge
of the human factors in planning and scheduling. In B.L. MacCarthy & J.R. Wilson
(Eds.), Human Performance in Planning and Scheduling: Fieldwork Studies,
Methodologies and Research Issues. London: Taylor & Francis.
Duchessi, P., and R.M. O' Keefe (1990). A Knowledge-Based Approach to Production
Planning. Journal of the Operational Research Society 41:5: 377-390.
Dutton, J. M. and W. Starbuck (1971). Finding Charlie’s Run-Time Estimator. In J.M. Dutton
and W. Starbuck (Eds.) Computer Simulation of Human Behaviour, Wiley. pp 218-242.

16
Dutton, J.M.(1962). Simulation of an Actual Production Scheduling and Workflow Control
System. International Journal of Production Research 4, 421-441.
Dutton, J.M.(1964). Production Scheduling: a Behaviour Model. International Journal of
Production Research 3, 3-27.
Easterby-Smith, M., Thorpe, R., & Lowe, A. (1991). Management Research – An
Introduction. London: Sage Publications.
Fox, P.D. and C.H. Kriebel (1967). An Empirical Study of Scheduling Decision Behavior.
The Journal of Industrial Engineering 18:6: 354-360.
Gary, K., Uzsoy, R., Smith, S.P., & Kempf, K.G.(1995). Measuring the quality of
manufacturing schedules. In D.E.Brown & W.T.Scherer (Eds.), Intelligent scheduling
systems (pp.129–154). Boston: Kluwer Academic Publishers.
Greene, J.H. (ed.) (1970). Production & Inventory Control Handbook. New York: McGraw-
Hill.
Grills, S. (ed.) (1998) Doing Ethnographic Research – Fieldwork Settings. London: Sage
Publications.
Gummesson, E. (1988). Qualitative Methods In Management Research. Sweden: Chartwell-
Bratt.
Hurst, E.G. and A. B. McNamara (1967). Heuristic Scheduling in a Woolen Mill.
Management Science 14:2:182-203.
Hutchins, E. (1995) Cognition In The Wild. Cambridge: MIT Press.
Jorgensen, D.L. (1989). Participant Observation: A Methodology for Human Studies, London:
Sage Publications.
Knoeppel, C.E. (1920). Graphic Production Control, New York: The Engineering Magazine.
Koepke, C. A. (1941). Plant Production Control, New York: Wiley & Sons.
Lopez, P., P. Esquirol, L. Haudot and M. Sicard (1998). Co-operative System Design in
Scheduling. International Journal of Production Research 36:1: 211-230.
Magee, J.F. (1958). Production Planning and Inventory Control, New York: McGraw-Hill.
Markus, L., & Tanis, C. (2000). The Enterprise Systems Experience - From Adoption to
Success. In Framing The Domains Of IT Management: Projecting the Future Through
the Past, Zmud, R.W. and M.F. Price (eds.). Cincinnati: Pinnaflex Educational
Resources, Inc., 173-207.
McKay K.N., Safayeni F., and Buzacott, J.A. (1995a). Common Sense Realities in Planning
and Scheduling Printed Circuit Board Production. International Journal of Production
Research, 33(6), 1587-1603.
McKay K.N., Safayeni F., and Buzacott, J.A. (1995b). Schedulers & Planners: What and How
Can We Learn From Them. In Brown D.E. and W.T. Scherer (eds), Intelligent
Scheduling Systems, (pp. 41-62), Boston: Kluwer Publishers.
McKay, K.N. & Buzacott, J.A. (2000). The application of computerized production control
systems in job shop environments. Computers in Industry, 42(2-3), 79–97.
McKay, K.N. (1992). Production Planning and Scheduling: A Model for Manufacturing
Decisions Requiring Judgement. PhD thesis, University of Waterloo, Ontario, Canada.
McKay, K.N., & Wiers, V.C.S. (2001). Decision support for production scheduling tasks in
shops with much uncertainty and little autonomous flexibility. In B.L. MacCarthy &
J.R. Wilson (Eds.), Human Performance in Planning and Scheduling: Fieldwork
Studies, Methodologies and Research Issues. London: Taylor & Francis, pp. 165-177.
McKay, K.N., & Wiers, V.C.S. (2003). Integrated Decision Support for Planning, Scheduling,
and Dispatching Tasks in a Focused Factory. Accepted for publication in Computers In
Industry.

17
 Norman, P. and S. Naveed (1990). A Comparison of Expert System and Human Operator
Performance for Cement Kiln Operation. Journal of the Operational Research Society
41:11: 1007- 1019.
Orlicky, J. (1975), Material Requirements Planning (London: McGraw-Hill).
Pinedo, M. (1995). Scheduling: theory, algorithms and systems. New York: Prentice Hall.
Pinedo, M. and Chao, X. (1999). Operations Scheduling With Applications in Manufacturing
and Services, Boston: Irwin McGraw-Hill.
Plossl, G.W. and Wight, O. (1967). Production and Inventory Control (S.I.: Prentice-Hall).
Sanderson, P. M. (1991). Towards the Model Human Scheduler. International Journal of
Human Factors in Manufacturing 1: 3:195-219.
Sanderson, P. M. and N. Moray (1990). The Human Factors of Scheduling Behaviour. In W.
Karwowski and M.Rahimi (eds.) Ergonomics of Hybrid Automated Systems II, Elsevier
pp.399-406.
Sanderson, P.M. (1989). The Human Planning and Scheduling Role in Advanced
Manufacturing Systems: An Emerging Human Factors Domain, Human Factors, 31 (6),
635–666.
Schensul, S.L., Schensul, J.J. and M.D. LeCompte (1999) Essential Ethnographic Methods.
New York: Altamira Press.
Spradley, J.P. (1979) The Ethnographic Interview. New York: Holt, Rinehart and Winston
Stadtler, H., & Kilger, C. (eds.) (2000). Supply chain management and advanced planning:
concepts, models, software and case studies. Berlin: Springer.
Stoop, P.P.M. (1996). Performance measurement in manufacturing: A method for short term
performance evaluation and diagnosis. Ph.D.Thesis Eindhoven University of
Technology.
Tabe, T., Yamamuro, S. and Salvendy, G. (1988). An Approach to Knowledge Elicitation in
Scheduling FMS: Toward a Hybrid Intelligent System. In W. Karwowski, H.R. Parsaei
and M.R. Wilhelm (eds.) Ergonomics of Hybrid Automated Systems I. Elsevier Science
Publishers, New-York, pp. 259-266.
Thurley, K.E.and A.C.Hamblin (1962). The Supervisor’s Role in Production Control.
International Journal of Production Research, 1, 1-12.
Vollmann, T.E., Berry, W.L., and Whybark, D.C.(1988). Manufacturing planning and control
systems. Homewood, IL: Irwin.
Wiers, V.C.S. (1997). Human–computer interaction in production scheduling: Analysis and
design of decision support systems for production scheduling tasks. Ph.D. Thesis
Eindhoven University of Technology
Wiers, V.C.S. (2001). Design of a Knowledge Based Scheduling System for a Sheet Material
Manufacturer. In B.L. MacCarthy & J.R. Wilson (Eds.), Human Performance in
Planning and Scheduling: Fieldwork Studies, Methodologies and Research Issues.
London: Taylor & Francis.
Wiers, V.C.S. (2002). A case study on the integration of APS and ERP in a steel processing
plant. Production Planning & Control 13(6), 552-560.
Wiers, V.C.S., and Schaaf, T.W. van der, (1997). A Framework for Decision Support in
Production Scheduling Tasks. Production Planning and Control, 8(6), 533–544.
Wortmann, J. C. (1992). Flexibility of Standard Software Packages for Production/Inventory
Control. Production Planning & Control, 3(3), 290-299.
Yin, R.K. (1989) Case Study Research: Design and Methods, London: Sage Publications.

18

View publication stats