human resources, scheduling,

loading, availability, simulation

Maria BARON-PUDA, Janusz MLECZKO*

SIMULATION OF HUMAN RESOURCES ALLOCATION

IN SCHEDULING PROCESSES

Abstract:
In the manufacturing planning process one of the main question is how many
and what kind of employees we need to complete successfully orders and jobs.
The best use of human resources at the shop floor level is one of the key issues.
The problem is complex especially in unit and small batch production systems,
where the changes in production program and production profile are much more
frequent than in large batch and mass production ones, so it needs the computer
support. This paper is focused on human resources planning at operational level,
which deals with day-to-day work. The paper shows how different methods of
scheduling influence human resource requirements.

1. RESOURCES IN COMPANY
For the correct production run a company needs some resources. They are especially
machinery and equipment, materials, financial resources, human resources (labour),
information technology and others.
One of the main resource’s attribute is its availability which is most often constrained by
many factors. Then, the resource capability is a key issue to determine its best use in the
production systems. In computerized management systems, the resources are divided as:
 un-renewable production resources, such as materials, energy – they cannot be remade or
regenerated to utilization,
 renewable production resources, as for example workstations, human resources, which can
be re-used again. The availability of employees as well as machines is constrained by the
labour calendar including working days, holidays and other breaks, by company’s
regulations defining daily number of working hours, number of shifts, length of breaks for
rest, lunch etc. and by number of similar resources (e.g. machines, employees).
In the production planning process, the main problem is to allocate resources to orders and
jobs in such way to maximize resource utilization and shop-floor productivity as well as
minimizing flow times, wastes and costs. From viewpoint of resource capacity utilization there
are possible three following situations:
– Overload – when the required resource capacity is more than available resource capacity.

*
PhD.; PhD, Bielsko-Biała University, Faculty of Mechanical Engineering and Computer Sciences,
Willowa 2, 43-309 Bielsko-Biała, Poland, [email protected], [email protected]

1
– Underload – when the required resource capacity is less than available resource capacity.
– Balance – required and available resource capacities are equal.
The full utilization of resources in the practice is very unlikely by reason of many
constraints and unforeseen circumstances. For example employee absenteeism and tardiness,
material shortage, machine failure are examples of day-to-day problems for managers.

2. PRODUCTION PLANNING AND SCHEDULING

Planning of resources can be made at one of the three levels – strategic, tactical and
operational ones, which refer to different planning period (longer or shorter time scale) and are
related to different levels of detailness (broader or narrower scope) [15].
Operational planning includes scheduling which aims at determining the optimal
manufacturing programs, load for workstations (machines) and setting reserves needed to
realize the production program in the required time span. In other words, scheduling comes
down to allocate available resources from viewpoint of time and area in the best possible way.
To built the optimal schedule during the satisfied time is usually impossible due to complex
nature of this problem and many constrains. The mathematical methods as well as techniques
of simulation and Constraint Theory may be applied to find the best solution. Examples of
solved problems for production planning and scheduling presented in the subject literature are
usually unsatisfied because of their sub-optimal, approximate or specific character. The
analyzed issues usually are narrowed and (or) they take into account scarce criteria. Therefore,
in the practice, the planning and scheduling tasks are based on engineer’s experience and
intuition. Enterprises improve their planning and scheduling by gathering the real data from
processes and creating the own knowledge data. [9]
In today’s companies planning and scheduling without informatics support means
practically impossible due to their complexity. Independently on the company’s manufacturing
system, the basic objective of the scheduling is to determine and calculate resources
requirement. Scheduling is used to assign jobs to machines in such way that processing time
and costs are minimized. There are also useful from viewpoint of planning human resources,
purchasing materials. The scheduling process is influenced by start date and due date, but there
are also many other variables and factors concerning it, for example kind of jobs to be
processed, resource availability including especially human resource capacities, processing
time, set-up time, planned maintenance, number of shifts etc.

2.1. Scheduling approaches

There are two classical approaches to scheduling: backward and forward. The both include
“what-if” scenarios. The forward scheduling determine jobs starting from the earliest possible
time, whereas the backward scheduling makes it in a reverse manner and determine jobs from
viewpoint of the latest possible time [12].
The backward scheduling answers a question “When the job should be stared at the latest in
order to finish it at the demanded date?”. This method is useful from the viewpoint of the
identification of bottlenecks. The manufacturing process should be started not later than in the
latest start date. Having the data on operations time, it is possible to determine the load for
particular machines (machine groups) used for manufacturing some products. The forward
scheduling answers a question “When the job will be finished if we start at the defined date?”.
Besides of backward scheduling and forward scheduling, it can be also used combined

2
scheduling approach [10]. First, the jobs are scheduled backward, but if a starting date for any
job is found out of the determined date (usually current date), this job is rescheduled moving
forward in time. This approach is called as midpoint scheduling.

2.2. Balancing

In today’s company’s practice there are usually some conflicts connected with resource
allocation, for example we have resources in excess as appropriate to demanded goals and jobs,
and conversely – we have resources’ shortages to do jobs by required time. To solve these
problems the load-capacity balancing function is used. It enables to compare resource capacity
demanded versus available resource capacities and check if a production orders and (or) jobs
are feasible or not by the required time and in consequence to make the best decision. The
capacity of resources may be considered as infinite or finite. Infinite approach ignores capacity
constraints, but can be helpful to identify bottlenecks in a proposed schedule.

Fig. 1. Analysis of availability for workstations regarding balancing function [9]

3. HUMAN RESOURCES IN THE PRODUCTION PLANNING

3.1. Human resource planning process

The human resource planning aims especially at providing the company with the right
people in the right place and at the right time and at motivating them to involvement and better
and better productivity. It can assist the organization to foresee changes and identify trends in
human resources and to adopt the personnel policies in order to avoid major problems [11].
At present, planning has a much shorter time-scale than in previous decades due to the
unpredictable changes into organizations as well as around them and the need to respond in
a flexible way. Moreover, there is more emphasis to develop multi-scenario plan that provides

3
flexibility to take advantage of opportunities that arise and to cope with unexpected dangers
that may befall [14].
Human resource planning process is divided into some steps which involve identifying
future needs and internal supply, and next, identifying the discrepancies between the two and
developing strategies to fill them – see Fig. 2.

Fig. 2. Stages in human resources planning

Labour demand deals with the human resources that will be needed to carry out the future
objectives and tasks from viewpoint of their number, competencies and locations. Labour
supply involves the current staff profile available within the organization.
The result of the comparison of labour demand and supply is the identification of gaps
(shortages or surpluses) in staffing needs [16]. A shortage indicates a future shortage of
required employees. The strategies to attract and develop staff with demanded competencies
will be needed. They could be for example training workers for new skills and knowledge,
overtime work. A surplus indicates the future excess in some categories of employees. In this
situation the strategies of dealing with excess in number of employees and (or) competencies
no longer needed in the organization should be developed. The example of strategies to address
surplus situations may be re-training, staff transfers, dismissal due to redundancy.
There are many different strategies the organization may use to address gaps, so the
decision which strategy, or which combination of strategies to choose, should be considered
carefully. During the process of analyzing the possible strategies the following factors should
be also considered [16]:
 Time – is there enough time to develop competencies of own staff internally or is special,
fast recruitment the best approach?

4
 Resources – the availability of demanded resources is also important as well as priorities
and timing.
 Cost – which strategy is of the lowest cost?

3.2. Planning the number and skills of workers demanded

From viewpoint of production process planning and control it’s very important to know
how many and what kind of employees are required in a given period. Knowing the quantity
and quality of needed resources we can estimate the time of order completion as well as the
labour costs related to employees involved in the manufacturing processes. The problem of
human resources planning is complex because human factor is the most unforeseeable. It is
very difficult to foresee future human behaviour, absenteeism. Additionally, the labour law
introduces many limitations related to way and time of human resources use.
There are many techniques used to calculate how many workers are demanded. One of
them is technique based on the balance of working time [3, 13]. To estimate the number of
required workers in the given period we need to know the data on available working time per
one employee. The data necessary to calculate the number of employees demanded in a given
period (Ne) for operating a given kind of workstation are following:
 number of workstations (machines, machine groups) of a given kind – n,
 labor hours required by one workstation in a given period – Tw,
 average available working hours per one employee in a given period – Te.
The average available working time per one employee (T e) is calculated considering
standards complied from the labour code, as well as the average employee absenteeism.
Absenteeism embraces average number of days when an employee is absent due to holidays,
sickness and other reasons. Holidays are defined by labour law and labour calendar. On the
contrary, the sickness absence is estimated on the basis of the company’s statistical data. For
example, an average employee is absent due to holidays – 20 days per year, due to ill-health
about 10 days per year and 5 days due to other reasons. Total annual absence rate for this
example is 35 days.
The formula for calculating the average available working time per one employee in the
given planning period is following:
Te = number of whole weeks1 within a planning period  40 hours per week
+ number of working days beyond whole weeks  8 hours per day
– number of holidays within whole weeks (holidays  Sundays)  8 hours per day
– average absenteeism rate  8 hours per day.

Having such data we can find Ne as follows:

n  Tw
Ne  (1)
Te
Planning the quality of workers demanded we have to know what are the main
requirements of workstations, that means what knowledge and skills are needed, what is the
responsibility and so on. Such data enable to define the profile of a preferred worker.

1
The first whole week includes period from 1st day of a given planning period to 7th day, the second –
from 8th to 14th and so on.

5
 In determining work requirements the job evaluation may be helpful. It is a process of
analysis and assessment of a work content form viewpoint of demanded knowledge and skills,
scope of responsibility, effort and working conditions. Job evaluation is a valuable tool used
especially for basic pay differentiation purpose, but it can also be used in other areas, like
defining the required professional qualifications, identifying training needs etc. [4]. There are a
number of different job evaluation methods. Each analytical method consists of a set of factors
defining main qualification requirements.
The present production systems want flexible and agile workers, who can be shifted
dynamically to where they are needed [6]. The functional flexibility is one of the main types of
labor flexibility within Atkinson’s concept of flexible firm [2]. From cost point of view there is
a question if the functional flexibility is really effective for a company. On one hand
polyfunctional operators are more costly because higher qualifications should be compensated
by higher payment, but on the other, they allow the company to reduce a number of employees,
so in consequence this solution seems as more cost-effective. There is assumed that the labor
flexibility is cost-effective, but there are a little evidence confirming this assumption [15].
However, today’s enterprises seek qualified, motivated and open for changes operators.

3.3. The use of human resources at shop floor

One of the key aims of production planning at operational level is the best use of human
resources at shop floor. Human resources optimization involves allocation of employees to
machines and jobs to ensure that minimal number of engaged operators as well as the balance
between over and under staffing. That means the working time of an employee will be used in
the best way, without the overtime and idle time.
Availability of human resources is constrained by many factors – Fig. 3, such as: number of
operators, their skills, maximal working time per a day and week which is limited by labour
law, work organization including number of shifts, breaks within a working day.

Labour calendar

Work organization

Availability of human
Labour law
resources is
constrained by Number of employees

Skills deficiency

Financial budget

Fig. 3. Factors constrained the use of human resources

Capacity and utilization of human resources may be improved by multi-skilling. Multi-

skilling enables an employee to work flexibly on a variety of tasks [1]. In today’s
manufacturing systems, especially of unit and small batch production, operational flexibility
and mobility between vocations and workstations is required.

6
4. SIMULATION-BASED HUMAN RESOURCES ALLOCATION
To find the best solution on human resources use in the production scheduling, the
simulation method is helpful. Simulation as a part of decision support systems for production
scheduling provides a way to get detailed information about the consequences of scheduling
decisions [8]. There are a number of reasons for using simulation versus analytical models, but
the main usefulness of simulation is to identify the potential problems and prevent them before
the schedule will be introduced at the shop floor [8].
The simulation method may be used in order to analyse what are the human needs taking
into account different scheduling method and answer the questions: “How many operators and
what kind of skills are required from viewpoint of a given scheduling method?” and “What
scheduling method does enable to use the available human resources in the optimal way?”. In
subject literature related to computer integrated manufacturing management, employees are
considered as the secondary resource with the less meaning for the decision process [9]. This
approach is very simplified. Confining the role of human resources in decision making on
a given solution may lead to higher labour costs and finally higher production costs. The role
of employment is one of the key factors and it should be taken into consideration together with
other resources, i.e. machines and workstations.
As shown in Fig. 4 the proposed simulation-based human resources allocation system
includes three sub-systems:
 ERP system, which is necessary to download the operative data and generate production
schedules,
 sub-system for defining optimization procedures,
 sub-system for preparing and conducting simulation processes.

Fig. 4. Simulation-based HR allocation system

Simulation-based HR allocation system may support to optimize human resources. HR

optimization involves allocation of employees to machines and jobs to ensure the minimal
number of workers and balance between over and under staffing. The algorithm presented in
Fig. 5 includes stages of simulation of human resources allocation based on accepted criteria
and scheduling methods.

7
 Fig. 5. Algorithm of simulation based HR allocation

Having the production plan, loading for machine groups, operating rules (including shift
number, working time, breaks, absenteeism rate), the human resources allocation may be
developed based on various optimization criteria, such as: minimal number of workers,
maximal use of employee’ working time, maximal use of employees’ skills, balance between
over and under staffing. Then, the target function is:

Y = f(N, S, Tl, Ta) (2)

where:
N – number of required workers,
S – worker’s skill range (the possibility to operate different machines by a worker),
Tl – time of loading for a worker,
Ta – time of availability of an average worker.

8
 From viewpoint of optimization of labour time and costs, the minimal number of operators
is demanded as well as the best use of their working time without over time work and idle time.
In other words, the absolute difference between loading time for a worker and his (her)
availability time should be minimal:
N
 Tli  Ta  min (3)
i 1

where:
n = 1,2,…,N – employees and N tends towards the minimum

5. SIMULATION OF HUMAN RESOURCE DEMANDS IN DIFFERENT

SCHEDULING APPROACHS - example
5.1. The input data

There is a medium production company employing about 80 shop floor workers. The
products are manufactured in unit and small batches and that’s why the machines’ loading as
well as required workers’ numbers are changeable. The example for human resources
simulation is concerned to the Machining Department. It is conducted in order to simulate how
different scheduling methods influence quantitative and qualitative requirements for operators.
The analysis is made for 4 weeks’ planning period (from the 1 st to 28th of January 2008).
The ERP integrated system is used to download the operative date, generate orders in the
planning period and allocate them to machines. The part of the production plan is presented in
Fig. 6.

Fig. 6. Production plan for the planning period – an example

Orders are then divided into jobs and jobs are assigned to machine groups, so we can find
out how the particular machines are loaded. The loading may vary depending on used

9
scheduling method. Changes in loading influence changes in required number and skills of
workers.
In the manufacturing process different machines are engaged. The further analysis is
narrowed only to the milling machines marked in Fig. 7 as F01 to F08 machines.

Fig. 7. Machine-hours for milling machines in particular weeks form viewpoint of forward
scheduling approach

5.2. Simulation of quantitative human resources needs

The considered milling machines (millers) are divided into four sub-groups regarding their
different qualification requirements to operator:
 CNC millers (include F01, F02, F03, F04 machines),
 CNC large millers (F05, F06),
 CNC millers for rough machining (F07),
 manual millers (F08).
The required number of operators (N) for particular machine sub-groups is calculated
according to the following formula:

totalmachine hours required in the given period

N (4)
labour hours available by one operatorin the given period

Total machine-hours for the specified machine sub-groups are related to the loading of
milling machines. The loadings of particular machines in successive weeks of the planning
period are determined considering different scheduling methods. Basing on the data derived
form ERP system the machine-hours for considered machine sub-groups taking into account
the forward schedule are presented in Fig. 8.
To calculate how many operators are needed in the following weeks, the time of availability
of an average person is required to know. In considered example the labour hours available by
one operator are calculated considering the planning period. From 1/1/2008 to 28/1/2008 there

10
are four weeks and one holiday (1st of January). A working day per an operator lasts 8 hours.
Excluding 30 minutes break, each operator works 7,5 hours a day (37,5 hours a week).

800

700

machine-hours 600

500 CNC
CNC_large
400
CNC_rough
300 manual
200

100

0
01/2008 02/2008 03/2008 04/2008
w eeks of the planning period

Fig. 8. Loading of milling machine sub-group in the planning period for forward schedule

Considering above assumed, the required number of workers for CNC machines in the first
week (01/2008), considering the forward schedule, is:

444,48 hs.
N   11,85  12 operators
37,5 hs.
The scores for further weeks for CNC miller as well as other machine sub-groups are
included in the Tab. 1. All results are up rounded to full time employment.

Tab. 1. Required and available human resources in forward scheduling

Milling Required number of operators in successive weeks Available2 number of

machine operators in every
sub-groups 01/2008 02/2008 03/2008 04/2008 week
CNC 12 18 12 7 13
CNC_large 4 5 2 1 3
CNC_rough 3 4 1 1 3
Manual 5 7 6 6 6
Total 24 34 21 15 25

2
In presented example the available operators’ number is assumed as average in the scale of the whole
planning period.

11
 The same calculation are made taking into account backward and midpoint scheduling
methods.
Next step is to simulate the balance between under and over staffing. It consists in
comparing demanded versus supplied (available) operators’ numbers and defining the gas as
surpluses or shortages considering different scheduling approaches. The comparison may be
done for particular machine sub-groups as well as for the whole machine group.
The surplus is when available number of operator is more than required (marked as “+”),
the shortage – when it is less (marked as “–“). The figures form 9 to 11 show the gaps in
human resources for machine sub-groups as well as for all milling machines (total) in
considered period.

10
gap (surplus or shortage)

5
CNC
CNC_large
0 CNC_rough
Manual
Total
-5

01/2008 02/2008 03/2008 04/2008

-10

Fig. 9. Surpluses or shortages in human resources in forward scheduling

10
gap (surplus or shortage)

5
CNC
CNC_large
CNC_rough
0
Manual
Total
-5

01/2008 02/2008 03/2008 04/2008

-10

Fig. 10. Surpluses or shortages in human resources in backward scheduling

12
 10

gap (surplus or shortage)

5
CNC
CNC_large
0 CNC_rough
Manual
Total
-5

01/2008 02/2008 03/2008 04/2008

-10

Fig. 11. Surpluses or shortages in human resources in midpoint scheduling

The quantitative needs should be next completed by defining qualifications requirements.

5.3. The quality of the required operators

To find out what are the main requirements of workstations (machine sub-groups)
addressed to job-holders, the job evaluation process is needed. The job evaluation makes
possible to define what are the main requirements of particular machines from viewpoint of
demanded qualifications (preferred knowledge and skills, scope of responsibility, effort etc.).

Milling machines
Sub-factor CNC CNC_large CNC_rough Manual
Education and experience required 40 30 30 20
Additional skills 1 3 1 0
Dexterity 1 4 4 5
Complexity of work 15 9 9 10
Responsibility for own work 10 12 10 6
Responsibility for others' work 0 0 0 0
Responsibility for external contacts 0 0 0 0
Mental effort 15 13 10 8
Physical effort 6 10 9 13
Noxiousness of worktime schedules 0 0 0 0
Monotony at work 2 4 4 2
Material environment conditions 2 3 2 4
Total score: 92 88 79 68

Fig. 12. Job evaluation assessment sheet

13
 The job evaluation exercise has been done using the method which is the most appropriate
in respect of the kind of workstations in the company. During job evaluation process every
workstation is considered according to specific factors and allocated in appropriate number of
points related to a pre-defined scale. The same factors and point scales are used for every
workstation. Each factor is broken down into levels of difficulty. The level is described in words
and in appropriate number of points. The higher level the more demanding factors and in
consequence the greater number of points. The results of job evaluation conducted for considered
milling machines are presented in Fig. 12.
The final result of job evaluation is to group workstations into some grades with like level
of qualification requirements – Tab. 2.

Tab. 2. Classification of machines into grades

Grade Point range Workstations

I (… – 70) Manual Miller
II (71 – 85) CNC miller (for rough machining)
III (86 – 100) CNC large miller, CNC miller
IV (101 – 115) …

Successively, job-holders are classified to appropriate grades. A multi-skilled worker able

to operate different machines from different grades is classified to the highest grade, according
the formula:
Gej = max{Gwi} (5)
where:
Gej – employee’s grade,
Gwi – workstation’s grade,
i= 1,2,…n – workstations which a j-employee can operate.

Table 2, as a result of the job evaluation, illustrates positions of particular workstations in

their relationships with others according to their different demands for knowledge and skills,
responsibility, effort, working conditions. This is next the fundamental for basic pay rate
differentiation and finally, for simulation of the labour costs related to considered production
plan.

5.4. Discussion of result

The simulation of three different cases of human resources demands has been made. Each
simulation result is related with some gaps.
From viewpoint of the forward scheduling:
 the greatest number of employees engaged to particular machines as well as to all milling
machines is needed,
 in the most critical week (02/2008) the human resources requirements are largest (34
operators in total are needed),
 the first week introduces 1 operator in excess,

14
 there are the greatest gaps (surpluses and shortages) in particular weeks (for ex. in the week
02/2008 9 workers are lacking and in turn, in the week 04/2008 10 employees are in
excess),
 shortages in required staff in the second week may be balanced by surplus in 04/2008 week.
The backward scheduling:
 requires the smallest operators’ number,
 in the starting week the shortages in certain machine sub-groups may be balanced by
surpluses in others, so in the global sense there would be no gap,
 in the second week the staff requirements are covered with its availability in three machine
sub-groups,
 there is relatively large surplus in human resources in the last week,
 generates the best matching between staff supply and demand for manual milling machines
in particular weeks.
In the scheduling with midpoint approach:
 the first week starts with 2 operators in excess,
 there is relatively great deficit in needed staff in the week 02/2008 (7 operators are lacking
in total), but this may be compensated by equivalent surpluses in the third and forth weeks,
 the surplus in the last week in the smallest,
 the number of demanded workers is generally more than in backward scheduling but less
than in forward one.

In conclusion, comparing the available versus demanded human resources in numerical

aspect for the assumed production plan, the backward scheduling approach is the best. It
enables to realize orders in time with the minimum number of operators. The workers are used
in the best way, that means numbers of available as well as demanded employees are consistent
in most of cases (considering weeks and machine sub-groups). Additionally, introducing the
adequate multi-skilling level where some workers would be able to operate different machines
makes possible to improve the organization of working day (less inactivity time, less over-time
work).

6. CONCLUSIONS
To realize the production plan in time, the required resources should be supplied. The
human resources are one of them and they are very important because human factor is one of
the most unforeseeable. There are many constraints, which make the time and way of human
resources use limited.
In unit and small batch production the demand for number and skills of employees varies
over time. In such environment employees are required to be more mobile and flexible between
workstations as well as between vocations, so workers should be encouraged to develop
multiple skills. To determine a number and quality of human resources required in the planning
period, the support of computer integrated system is necessary. ERP system enable to have
actual and integrated information on the company’s state and condition as well as it gives the
data to perform “what-if” analysis and see the impact of different changes in order to avoid
making costly decisions.
With the use of ERP system we can simulate how different scheduling methods influence
human resources needs, that means how many employees and what skills are required. These

15
information may be useful in making decision on the best use of human resources from time
and cost point of view.

References
[1] ARMSTRONG M.: A Handbook of Human Resource Management Practice. Kogan Page,
London 2003.
[2] ATKINSON J.: Manpower Strategies for Flexible Organizations. “Personnel Manage-
ment”, August 1984, pp. 28-31.
[3] BARON-PUDA M.: Personnel Management. Wydawnictwo Akademii Techniczno-
Humanistycznej, Bielsko-Biała 2006.
[4] BARON-PUDA M.: Chosen problems of personnel management in industrial enterprises
(2).Benefits of job evaluation. Produktywność i Innowacje, nr 1/2006, pp. 43-48.
[5] HANNAGAN T.: Management. Concepts and Practices (3/e). Pearson Education Ltd.,
Harlow 2002.
[6] HOPP W.J., VAN OYEN M.P.: Agile Workforce Evaluation: A Framework for Cross.
Training and Coordination, IIE Transactions, Volume 36, no.10, October 2004, pp. 919-
940.
[7] HUNSAKER P.L.: Management. A Skill Approach (2/e). Prentice Hall, New York 2004
[8] LEHTONEN, J.-M.; APPELQVIST, P.; RUOHOLA, T.; MATTILA: Simulation-based
finite scheduling at Albany International. Proceedings of the 2003 Winter Simulation
Conference, Volume 2, 7-10 Dec. 2003, pp. 1449 - 1455.
[9] MLECZKO J.: Komputerowo wspomagane zarządzanie. Wydawnictwo Centrum Nowych
Technologii, Bielsko-Biała 2008.
[10] MLECZKO J.: Komputerowo wspomagane planowania przebiegów procesów
produkcyjnych. Wydawnictwo Centrum Nowych Technologii, Bielsko-Biała 2008.
[11] MULLINS L.J.: Management and Organizational Behaviour (6/e). FT Prentice Hall,
London 2002.
[12] OGUZ C., FIKRET ERCAN M., EDWIN CHENG T. C. and FUNG Y. F.: Heuristic
algorithms for multiprocessor task scheduling in a two-stage hybrid flow-shop. European
Journal of Operational Research, Volume 149, Issue 2, 1 September 2003, pp. 390-403.
[13] SEKUŁA Z.: Planowanie zatrudnienia. Oficyna Ekonomiczna, Kraków 2001.
[14] STREDWICK J.: An Introduction to Human Resource Management. Butterworth
Heinemann, Oxford 2000.
[15] VALVERDE M., TREGASKIS O. and BREWSTER C.: Labor flexibility and firm
performance. International Advances in Economic Research. Volume 6, No. 4, November
2000, pp. 649-66.
[16] Workforce Planning Guide. www.hr.state.tx.us

16