Conference PaperPDF Available

Worker Assignment Problems in Manufacturing Systems: a Literature Analysis

Authors:

Abstract and Figures

Workers assignment problems (WAPs) are found in several types of manufacturing systems. Typically, WAPs are known to greatly impact the system performance. Numerous research works related to WAPs have been published. In this work, we analyze this literature according to several types of criteria. We first take into account such important features of the WAP as workers flexibility, the number of workers with respect to the number of stations, workers movement, the problem nature (static or dynamic assignment) and human consideration (e. g. learning and forgetting effects). We also take into consideration the type of system it is related to (e. g. U-shaped systems, and cellular organizations). The different methods used to solve these problems are distinguished (e. g. mathematical approaches, simulation, hybrid methods, etc.). Several research directions are pointed out and discussed in the conclusion. © 2013 International Institute for Innovation, Industrial Engineering and Entrepreneurship - I4e2.
Content may be subject to copyright.
Workers Assignment Problems in
Manufacturing Systems: a Literature Analysis
(presented at the 5th IESM Conference, October 2013, Rabat, Morocco) © I4e2 2013
Achraf Ammar
LOGIQ Research group,
University of Sfax, Tunisia
and LIMOS, UMR CNRS 6158,
Clermont University,
Aubière France
achraf.ammar@ifma.fr
Henri Pierreval
LIMOS, UMR CNRS 6158,
Clermont University
Aubière, France
henri.pierreval@ifma.fr
Sabeur Elkosentini
LOGIQ Research group,
University of Sfax
Sfax, Tunisia
Sabeur.Elkosantini@isima.rnu.tn
Abstract— Workers assignment problems (WAPs) are
found in several types of manufacturing systems. Typically,
WAPs are known to greatly impact the system performance.
Numerous research works related to WAPs have been
published. In this work, we analyze this literature according
to several types of criteria. We first take into account such
important features of the WAP as workers flexibility, the
number of workers with respect to the number of stations,
workers movement, the problem nature (static or dynamic
assignment) and human consideration (e. g. learning and
forgetting effects). We also take into consideration the type
of system it is related to (e. g. U-shaped systems, and cellular
organizations). The different methods used to solve these
problems are distinguished (e. g. mathematical approaches,
simulation, hybrid methods, etc.). Several research
directions are pointed out and discussed in the conclusion.
Keywords— worker assignment, dynamic assignment,
manufacturing systems, optimization methods, Simulation
I. INTRODUCTION
Workers assignment problems (WAPs) are among the
major problems that manufacturing companies have to
face. Indeed, the way workers are assigned can greatly
affect the system performance. Due to the variety of
WAPs, there exist various definitions, assumptions and
methods related to these problems. Often, WAPs are
concerned with specific contexts, such as assembly line
balancing and cell formation problems. These problems
are known for their complexity and they are often NP-Hard
[1, 2]. Many articles have been published that aim at
addressing various types of problems related to human
resources. Part of this literature has been analyzed by
researchers. Bidanda et al. [3] analyze diverse types of
human issues involved in a particular type of system,
which are cellular manufacturing systems. An old survey
conducted in 1976 about workforce allocation in cyclical
scheduling problems is presented by Baker [4]. Much
more recently, Bergh et al [5] focus on how issues related
to personal are taken into account in the scheduling of
several types of systems (military, services, transportation,
manufacturing, etc.).
Although these works present certain common interests
with our literature analysis, they address different aspects.
In this article, emphasis is placed on workers assignment
in manufacturing systems. Therefore, we do not consider
other problems than assignment (e. g., other types of
human problems that arise to personnel or scheduling
problems that can involve workforce). We particularly
focus on individual workers, rather than on team of
workers (e. g. crew) and do not restrict ourselves to
cellular manufacturing. This leads us to consider several
important issues that are not analyzed in a scheduling
context (e. g. balancing problems, U-Line, walking
workers system, dual resource constrained, assignment
rules, etc.) but are important in the specific area of
manufacturing systems. Although this conference article
cannot be exhaustive, it relies on a large number of
literature references. We first focus, in Section 2, on
several general characteristics that define WAPs. Then, in
Section 3 we distinguish the specific types of
manufacturing systems that are concerned, in other words
the application contexts. Section 4 discusses the ways
WAPs are addressed, and the approaches that are used to
solve them. Finally, several research perspectives are
pointed out.
II. WORKERS ASSIGNMENT PROBLEMS IN
MANUFACTURING SYSTEMS : GENERAL FEATURES
In many manufacturing systems, if workers are costly,
they are also flexible resources that can be assigned to one
or several tasks (e. g. machining, transportation,
inspection, etc.), for a given period of time, to meet certain
given requirements (e. g., complete a production plan,
etc.). The set of available workers can be of a fixed size, or
can vary if the number of required workers is also taken
into account. In certain types of manufacturing systems,
the number of workers can be greater than the number of
operations or machines, as in labor intensive systems.
Thus, workers have to be assigned so as to perform
different operations on the same and/or different work
centers, in different periods [6]. On the contrary, other
systems have fewer workers than machines, as in machine
intensive systems or dual resources constrained (DRC)
systems where workers can be transferred across various
workstations as required [6, 7]. WAPs often aim at
optimizing one or several performance criteria such as
minimizing the cycle time, the lead times, the time to walk
from a station to another, the tardiness, the production and
training cost, the level of works in process, or the system
balance [3, 8].
A first important factor, which is taken into account in
many studies, is the workers flexibility, if they are trained
to perform several types of tasks. This allows the
possibility to assign them to different operations,
depending on their skill levels and on the changes
occurring in the production system. Workers are
considered as homogeneous, if they have the same skill
level, or heterogeneous, if they have different skill levels
(see for example [9]). Depending on the problem, it can be
more efficient to have homogenous workers or
heterogeneous workers [10].
Flexible workers can walk between workstations and
perform the product completely from the beginning to the
end, or partially in so called walking workers system
(WWS) [1, 11]. In such systems, the number of assigned
workers is adjusted in response to the production
requirement and the manpower cost is expected to
decrease. Workers may have different individual skills or
varying working speed when performing a task which
affects the efficiency of WWS [11]. However, according to
[1], the walking time is a factor that seems sometimes
neglected in some papers dealing with WWS.
Whether the work given to workers can be
collaborative or non collaborative is, also, an important
distinction in WAPs. In non collaborative systems only
one worker is assigned to perform a single task at a time.
In collaborative systems, more than one flexible worker
can be assigned to a job simultaneously. Most researchers
often assume that there is no collaboration between
workers [12].
Most articles dealing with WAPs address static
problems. In such cases, the production system features are
defined by a fixed number of workers, each machine
capacity is known and constant. Workers are generally
assigned for a single time period with deterministic and
constant product mix and demand and unexpected events
are not considered. As a consequence, such problems are
generally solved using classical optimization methods
(either exact, such as branch and bound, or inexact such as
heuristic methods). When many changes occur in the
system state, such as the arrival of new products or orders,
machine failures, workers unavailability, etc. and/or
stochastic considerations need to be taken into account,
then static approaches can no longer be relevant. This
requires changes in real time of the workers assignment. In
such cases the assignment of workers is said dynamic (see
[13] and [14] for example). This often appears to be a
relevant approach when bottlenecks may vary among
workstations.
In addition to the technical considerations we have
introduced, certain human aspects may be taken into
account in WAPs. They include the ability of worker to
acquire news skills, learning and forgetting phenomena
(see for example [15, 16 and 17]) and others human
behavioral such as fatigue, motivation, stress, etc. [18].
III. APPLICATION PROBLEMS
In the literature, WAPs are often studied in relation
with specific industrial problems, with the objective of
improving different performance measures. It is interesting
to distinguish these works according to the type of
industrial problem addressed.
A. Workers Assignment In Production Lines
a) Assembly Line Balancing Problems
In the scientific literature, the Simple Assembly Line
Balancing Problem (SALBP) is widely studied. Workers
are fixed assigned, operation times are deterministic and
cycle time is fixed and independent of workers skills (see
[19]). Miralles et al. [20, 21] introduce a workers
assignment in the assembly line based on the SALBP
concept. This problem, called the Assembly Line Worker
Assignment and Balancing Problem (ALWABP), consists
in providing a simultaneous solution to a double
assignment: tasks to stations and available workers to
stations, with the objectives of minimizing the overall
cycle time. The main novelty introduced by the ALWABP
concerns principally the operation times that may vary
from one worker to another. Moreover, the number of
workers per station may vary from one station to another
and the positions of the workers in the line become a
decision to be taken. The workers assignment is considered
as static for both the SALBP and ALWABP.
Recently, numerous research works have considered
workers assignment in more realistic lines, including
mixed-model or multi-model lines and stochastic times [8].
Flexible workers can be assigned dynamically into several
stations, move between them and collaborate to balance
the system simultaneously. Such problems are known as
Generalized Assembly Line Balancing Problem (GALBP).
Sennot et al. [12], Ahn et al. [22], McClain et al., [18],
Askin and Chen [23], Chen and Askin [24], Bartholdi et al.
[25, 26], Bratcu and Dolgui [27] Wang et al. [11] are
among the researchers who have addressed this problem.
Sennot et al. [12] study a non collaborative serial line by
assigning one specialist worker, i.e. which can perform
only one type of tasks, to each station and one floater
worker, i.e. fully skilled worker which can move to
perform any task on any machine, periodically to the
bottleneck station of the line. They demonstrate the
importance of assigning floater workers to balance the line
and reduce the work in process (WIP). Ahn et al. [22]
consider both collaboration and non-collaboration when
assigning dynamically two flexible workers to a two-
station line with random arrival and operating times. Their
objective was to minimize the holding cost. McClain et al.,
[18], Askin and Chen [23] and Chen and Askin [24]
consider systems where each worker is assigned to one
station to perform a fixed task, and each two successive
workers can perform a shared task. They are interested in
determining whether a worker who has just finished a
fixed task for a given job should pass the job down to the
next worker to do the shared task, or simply do it himself.
Bartholdi et al. [25, 26] and Bratcu and Dolgui [27] use
bucket brigade concepts to assign workers in an assembly
line. All workers are fully skilled, i.e. each of them can
perform any task, and be assigned on the line depending on
how fast he/she can perform tasks. When a downstream
worker completes a job or is blocked by a further
downstream worker, he/she moves back up the line and
takes over the task of the next upstream worker. The main
benefit reported when using this ‘bucket brigade’ method
is the self balancing of the line based on the movement of
workers. Wang et al. [11] also study the movement of fully
skilled workers so as to reduce the effect of bottlenecks in
serial assembly line. Each worker moves along the line to
carry out the tasks at each station. The aim of their work is
to minimize the waiting time that may affect the overall
system performance.
b) Worker Assignment in U-shaped Line
Several publications in the literature are concerned
with workers assignment in U-shaped lines. The traditional
straight line is here adapted to a U-shaped, and flexible
workers will be assigned to more than one station and will
move between the two legs of the U-line, to perform
combinations of tasks that would not be allowed in a
straight line layout. U-line allows the adjustment of the
number of workers. Figure 1 represents how five workers
in straight line are reduced to four only in U-line to
balance the line.
Most of the articles dealing with workers assignment in
U-line consider the static assignment of homogeneous
workers. Sparling and Miltenburg [27] analyze the static
assignment of workers in a mixed-model U-line by
determining the assignment of workers to stations under
predetermined cycle time. Nakade and Ohno [29] and
Nakade and Nishiwaki [9] study the static assignment of
workers in U-line to minimize the overall cycle time.
Nakade and Ohno [29] consider homogeneous workers
with stochastic walking and operating times, whereas
Nakade and Nishiwaki [9] consider heterogeneous workers
and deterministic operating and walking times.
Zavadlev et al. [30] study the dynamic assignment of
workers in a U-line using a strategy where each worker,
after the completion of an operation, moves backward or
forward and searches for an idle machine. They have
shown that this assignment approach is more effective than
the fixed assignment to balance the line and reduce WIP
inventory. Celano et al. [31] address the assignment of
flexible workers in mixed model U-line with the objective
of minimizing the total conveyor stoppage time.
Collaboration between workers is allowed. Each worker is
assigned to one station and can move forward or backward
to the adjacent station to help if necessarily the critical
worker who stopped the conveyor. Different workstations
lengths, worker walking times and speed are taken into
account to decide which available worker will be assigned
to help the worker in a critical station.
Fig 1: Workers assignment in straight and U-line
B. Workers Assignment and Cell Formation Problems
The cell formation problem (CFP) is the most
addressed problem in the design of cellular manufacturing
systems (CMS). Most research works dealing with workers
assignment in CMS focus on the static assignment of
workers into one or several cells after their formation,
considering homogeneous workers [32]. The processing
time of the job in the cell depends on the workers
assignment and system balancing is achieved by workers
movement between stations (intra-cell movement or inter-
cell movement) [6]. The assignment and the number of
workers can vary from one period to the next in order to
cope with a varying demand [33].
Other research works study heterogeneous workers.
Norman et al. [34] address the static assignment of
workers within a single cell taking into account technical
and human skills, in order to maximize the effectiveness of
the system, where effectiveness is assumed to be a
function of productivity, output quality, and training costs
associated with a particular worker assignment.
In the design of a CMS, several problems consider that
the production plan will change over time. In such cases,
the entire planning horizon is divided into time periods,
according to the differences in product mix and/or demand
in each period. Consequently, the formed cells and the
assignment of workers to machines or cells in a given
period may not be suited for the next period. Machines can
be relocated between the cells, removed or added. Then the
assignment of workers can be defined simultaneously with
the cells, using cross training and hiring or firing. Min and
Shin [35] state that CMS will not be successful unless,
cells are formed and workers are assigned simultaneously.
Aryanezhad et al. [36] and Mahdavi et al. [2] study
simultaneously cell formation and assignment of flexible
workers with respect to multi-period production planning,
with the objective of minimizing the costs induced by
facilities and humans (production cost, material handling,
hiring, firing, training cost, etc). These authors consider
deterministic production requirements and static
assignment of workers for each period.
According to [13], very little research considers the
dynamic assignment of workers in CMS. Cesani and
Steudel [13] focus on dynamic workers assignment in
CMS characterized by intra-cell workers movement, a
number of workers less than the number of machines,
multiple products, random inter-arrivals times and arrivals
of jobs. When a worker completes a job at a machine,
he/she moves to work on the next job that has been waiting
the longest in any of his/her assigned workstations. These
authors compare three workers assignment strategies and
their impact on system balancing: dedicated (when only
one worker is responsible for a machine or group of
machines), shared (when more than one worker collaborate
for a machine or a group of machines) and combined
assignments (when the worker has both dedicated and
shared machine assignments). The main aspects
considered are related to the workload balancing and
sharing between workstations and workers and what to do
to the bottleneck workstations in order to keep the best
system performance.
C. Workers Assignment in Dual Resource Constrained
Job Shop
Dual Resource Constrained (DRC) systems are
characterized by a number of workers who are cross-
trained and flexible although they are subjected to learning
and forgetting phenomena [7]. There are fewer workers
than machines. In such environments, the assignment
decisions must consider both machine capacities and
workers availability [37]. Workers assignment in DRC job
shop systems is mainly addressed through assignment
rules, also known as ‘when’ and ‘where’ rules. When rules
are used to determine when a worker should be assigned to
another machine, and where rules are used to select the
machine to which the worker will be assigned. Xu et al. [7]
provided an overview of workers assignment rules and
other technical challenges to be considered in DRC job
shop systems. These assignment rules permit the
assignment of workers in real time.
Centralized (CEN) and decentralized (DEC) are the
two main types of ‘when’ assignment rules. The CEN rule
allows a worker to be assigned to another machine once a
job is completed, even if there are more jobs waiting in the
current machine’s queue. This rule increases the
movement of workers between workstations. It is used in
case where workers are heterogeneous. In contrast, with
the DEC rule, a worker is allowed to be assigned to
another machine only if the current machine’s queue has
been empty. Xu et al. [7] argue that when rules are more
important and have more effect on the overall system
performance than the where rules. Bobrowski and Park
[14] and Kher [37] describe and test different workers
assignment rules in DRC job shop environment with high
flexibility of workers and stochastic arrival and processing
time, so as to balance of the system in real time especially
in the presence of uncertainty. Araz and Salum [38] argue
that there is no study on DRC system in which the 'where'
and 'when' workers assignment rules are dynamically
selected to respond to new manufacturing conditions. They
propose a real-time assignment methodology for DRC
systems to address this problem.
IV. APPROACHES USED TO SOLVE WAPS
It is clear that the approaches used to solve WAPs
strongly depend on their characteristics, as explained in the
previous sections. Meanwhile, there is some material about
the solving strategies found in the literature that is generic
enough or that can be re-used or adapted from one WAP
case to another, which justifies it is being worth analyzed
in this section. The approaches used by researchers can
involve optimization, multi-criteria decision making, or
performance evaluation issues.
A. Mathematical Formulation
In numerous studies, WAPs are considered as an
optimization problem with either single or multiple
objectives, which can be linear or not. In some cases,
researchers take into consideration uncertain data and
stochastic formulations.
Several types of mathematical formulations have been
proposed for the WAPs with one single objective. A
representative one, used for the static workers assignment,
which can be adapted to a broader range of problems is the
one suggested in [39]. It is based on a mathematical
programming model that aims at minimizing the total cost
expanded on workers assignment to CMS. This
formulation is as follows.
s.t.
Where I represents the number of workers, J represents
the number of cells, K represents the number of machines
and L represents the number of products.
= Costs of worker i operating one unit of time
= 1 if worker i is assigned to operate on
machine k in cell j, 0 if not
= the technical merit of the worker i operating
machine k of a cells, , = 0 if worker
i is unable to operate on machine k.
= the quantity of product l to be manufactured in
a planning horizon.
= normalized time for manufacturing a product
l on the machine k of a cell j, and 0 if a product
cannot be manufactured by the machine k of a cell
j.
The objective function (1) represents the total cost of
all workers assigned to machines to produce the requested
quantity of products in a planning horizon. The first
constraint (equation (2)) ensures that each worker must be
assigned to only one machine of only one cell. Equation
(3) guaranties that each machine must be operated by only
one worker. Equation (4) restricts the assignment of
worker to machine on which he is able to operate. Miralles
et al. [20, 21] and Moreira and Costa [40] propose an
integer programming formulation to the ALWABP taking
into account the same types of constraints as stated above.
Suer [33] develops two assignment models for labor
intensive MC. A first mixed integer model is used to
assign workers into cell with the objective of maximizing
the production rate. The second model is used to ensure
that the demands are met with a minimum number of
workers using an integer programming formulation.
Aryanezhad et al. [36] propose a non linear integer
programming model to deal with a simultaneous dynamic
cell formation and workers assignment problem, which
they convert into a linear one. The objective function
consists of two separate components: machine-related
costs (production cost, inter-cell material handling cost,
machine costs in the planning horizon) and human costs
(hiring cost, firing cost, training cost and salary).
Gokhan and Suer [41] develop two non-linear
stochastic mathematical models to deal with worker
assignment and cell loading phases. In both models,
processing times and demand are probabilistic and
normally distributed. Nikoofarid and Aalaei [42] present a
mathematical formulation for a cell formation problem in a
dynamic cellular manufacturing system to manage
machines and workers assignment into cells over a certain
planning horizon.
Some other researchers have considered more than one
single objective for the WAPs. Sirovetnukul and Chutima
[1] present a multi objective model and a hierarchical
procedure to assign a minimum number of workers into a
U-line and minimize the deviation of operation and
walking times. Slomp et al. [43], and Mahdavi et al. [2]
develop a mathematical goal programming model for the
multi period CMS design and workers assignment.
Widyadana [44] addresses the WAPs in U-shaped
production line using goal programming for multi
objective problems with two goals: to minimize the
(1)
(2)
(3)
(4)
number of temporary workers to reduce cost and to
minimize the cycle time in each station. Min and Shin [35]
develop a multi-objective model to form manufacturing
cell and assign the workers simultaneously. The objectives
of the model are to maximize part similarities, match
workers skills to parts, minimize total machine processing
times, and minimize total labor costs. A sequential
heuristic that decomposes this problem into two smaller
problems: formation of the machine cells and assignment
of workers is proposed.
B. Optimization approaches
Many researchers address WAPs as combinatorial
optimization problems, which are generally NP-hard and
may require methods suited for large instances. Thus,
different approaches were developed to solve such
problems and to provide good or global or local optimum
solutions, given one or several performance objectives,
with respect to certain constraints. Miralles et al. [21]
apply a basic Branch and Bound approach with three
possible search strategies (depth first search with complete
node development, best first search and minimal lower
bound) and different parameters for the ALWABP. They
also suggest heuristics for larger problems.
More heuristic approaches have been used to find an
acceptable solution in reasonable time for several WAPs.
Blum and Miralles [45] present an iterative algorithm
based on beam search for the ALWABP with the objective
of minimizing the cycle time. Song et al. [46] propose a
recursive approach for assigning workers in assembly line
while considering the variance of workers skill. Nakade
and Ohno [29] and Nakade and Nishiwaki [9] propose an
algorithm to find the minimum number of workers to
assign to machines in a U-line, with the objective of
minimizing the overall cycle time. Campbell and Diaby
[47] develop a heuristic for assigning heterogeneous
workers to multiple work centers at the beginning of a
shift. They conclude that their heuristic outperforms a
lagrangian heuristic. Anzanello and Fogliatto [15] and
Thongnanit et al. [16] propose a method that uses learning
curves to guide the best assignment of product to workers
in assembly line.
Some researchers have applied meta-heuristics
algorithms to cope with the complexity of these problems.
Genetic algorithm (GA) seems to be the most used among
the meta-heuristics to address WAPs. ElMaraghy et al.
[48] and Elmaraghy et al. [49] use a GA approach for
workers assignment in DRC system. ElMaraghy et al. [49]
add the determination of the best number of workers and
machine and workers assignment to jobs along with
workers assignment rules. Celano et al. [31] use GA to
solve a workers assignment problem in manual mixed
model assembly U-line to minimize the total conveyor
stoppage time.
It seems that other types of meta-heuristic approaches
are much less used to solve this problem. Indeed, Tabu
Search (TS), Simulated Annealing (SA) and Ant Colony
(AC) and others approaches are seldom used alone to solve
WAPs. Instead, the effort is put on hybrid meta-heuristics,
which combine meta-heuristic with other optimization
techniques. Chaves et al. [50] adapt a hybrid heuristic
called Clustering Search (CS) to solve ALWABP. In the
CS, the evolutionary algorithm was substituted by distinct
meta-heuristics such as TS or SA. They use SA as meta-
heuristic to generate solutions to clustering process. Tao et
al. [51] use a Petri Net modeling method and a hybrid
genetic algorithm and simulated annealing algorithm
(GASA) search technique so as to deal with dynamic DRC
job shop and workers assignment. The proposed modeling
approach takes advantages of the search ability of GA and
the efficiency in avoiding a partial minimum of SA to
determine the new optimal assignment. The effectiveness
and feasibility of applying this proposed approach in
practical cases and even for real-time applications are
demonstrated by the authors.
C. Simulation modeling
Evaluating performances of systems where workers are
involved using simulation presents the advantage to avoid
unrealistic assumptions. DRC studies are probably those
that predominantly use simulation, due to their inherent
complexity and to the dynamic aspect of the problems
addressed. Bobrowski and Park [14], Kher [37] and Araz
and Salum [38] use simulation to evaluate several workers
assignment scenarios in DRC job shop system with
stochastic times. Bischak [52] uses simulation to compare
the throughput of U-shaped flexible workers lines with that
of serial lines with one fixed worker per machine. Ertay
and Ruan [6] use simulation to evaluate several
alternatives of workers assignment in CMS based on the
average lead-time of products and the average worker
utilization. These parameters are used as outputs criteria
for a Data Envelopment Analysis (DEA) approach that is
also applied to compare and evaluate the possible
alternatives of workers assignment in a CMS. Askin and
Chen [23] and Chen and Askin [24] develop a simulation
models to evaluate the performance of different decision
rules to dynamically assign tasks to workers in generalized
assembly line. Bartholdi et al.[25, 26] and Bratcu and
Dolgui [27] use simulation to analyze different instances
of workers assignment in Bucket brigades assembly line.
Shafer et al. [17] investigate through simulation how
patterns of learning and forgetting influence the
assignment of workers and affect the operating
performance in assembly line.
D. Combined approaches
Some works combine different approaches to deal with
WAPs. Simulation is often combined with other methods,
such as neural network, fuzzy approaches or optimization
algorithm. Araz and Salum [38] propose a multi-criteria
real-time methodology to dynamically select the worker
assignment rules for a DRC system. The methodology
employs artificial neural networks integrated with a
simulation model and fuzzy inference system in order to
reduce computing times and cope with multiple
performance criteria. Liu and Li [53] consider the WAP as
a fuzzy quadratic assignment problem with penalty
(FQAPP). A simulation and knowledge-based hybrid
genetic algorithm was designed to solve this problem.
Azadeh et al. [54] present an integrated method to select
the best scenario to assign workers in a CMS using a
simulation model and a fuzzy data envelopment analysis
(FDEA) approach. Azadeh et al. [55] use an integrated
simulation-genetic algorithm to determine the number of
cross trained workers to be assigned in a CMS. Azadeh et
al. [56] use an integrated fuzzy Multi Criteria Decision
Making (MCDM) approach based on fuzzy analysis
hierarchical process (AHP) and the Technique for Order
Performance by Similarity to Ideal Solution (TOPSIS) to
assign workers in CMS. The proposed methodology uses
triangular fuzzy numbers for AHP and computer
simulation. Linguistic variables and experts’ judgments are
used with the intention to make the decision making
process more realistic and reliable.
V. CONCLUSION
This article presents a literature review for WAPs in
manufacturing systems. A significant, but not exhaustive,
list of research papers was identified and analyzed, mainly
based on different aspects. We distinguished different
features that characterize WAPs (e. g., workers flexibility,
training possibilities, collaboration and movement of
workers, impact of human aspects and dynamic vs. static
assignment). We also distinguished the types of
manufacturing systems that are mostly studied by
researchers and the different methods used to address
WAPs, including optimization, simulation, multi-criteria
decision making and fuzzy theory or a combination of
these approaches.
Several observations can be drawn from our analysis,
which are useful for identifying research directions. Most
research works address WAPs on the basis of ‘technical’
considerations (e. g., workers skills, movement, etc). The
effect of more ‘human’ behavioral considerations (e. g.
fatigue) when assigning workers seems to be much less
studied. For example, taking into account such factors as
fatigue, stress, etc. would worth being addressed especially
in system where the operations are dangerous or difficult
to carry out. Let us also note that most researchers dealing
with workers flexibility take into account their ability to
perform tasks (allowed by their skills) through differences
in the time they need to perform different given operations.
However, others important criteria are impacted by the
worker’s skills, such as the quality of the jobs performed
(which can be for example measured through the
occurrence of defects). Since flexibility is obtained thanks
to the training of workers, assignment strategies can also
involve training objectives. It is the case if the assignment
is based on such principles as job rotation [45],
collaborative work and chaining flexibility [22]. Research
studies to better understand how performance and training
could be simultaneously achieved would be useful.
We have, also seen that, except for DRC job shop,
most articles deal with static problems. Therefore, taking
into account unexpected events and uncertain
environments through suited reactive assignments
strategies remains a relevant global research direction in
many cases. In addition, several types of problems would
need to be addressed with more realistic assumptions (for
example, the movement of flexible workers between
stations, is often considered to be negligible, machines are
assumed to be reliable, etc.).
We know that WAPs have an important impact on the
system performance. However, in many cases WAPs also
interact with other types of decisions. Typically, in a job
shop system, jobs priority has also a great effect.
Unfortunately, the dynamic assignment of workers and the
dynamic selection of dispatching rules are not studied
simultaneously, and research studies would be necessary to
better investigate such type of concurent decisions [38].
Regarding the approaches used for dynamic WAPs,
although neural network simulation metamodels were used
in [38], intelligent approaches, such as expert systems [57],
multi-agent approaches and learning methods [58] are little
reported, although they present an important potential.
This literature analysis and the research direction
pointed out show that the scientific area of workers
assignment problems remains quite opened to further
research.
REFERENCES
[1] R. Sirovetnukul and P. Chutima, “The impact of walking time on
U-shaped assembly line worker allocation problems,” Engineering
Journal, vol. 14, 2010.
[2] I. Mahdavi, A. Aalaei, M.M. Paydar and M. Solimanpur,
“Designing a mathematical model for dynamic cellular
manufacturing systems considering production planning and
worker assignment,” Computers and Mathematics with
Applications, vol. 60, 2010, pp. 1014–1025.
[3] B. Bidanda, P. Ariyawongrat, K.M. Needy, B.A. Norman and W.
Tharmmaphornphilas, “Human-related issues in manufacturing cell
design, implementation and operation: a review and survey,”
Computer & Industrial Engineering, vol. 48, 2005, 507–523.
[4] K.R. Baker, “Workforce allocation in cyclical scheduling
problems: a survey,” Operational Research Quarterly, vol. 27,
1976, pp. 155–167.
[5] J.V. Bergh, J. Belien, P.D. Bruecker, E. Demeulemeester and L.D.
Boeck, “Personnels scheduling:A literature review,” European
Journal of Operational Research, vol. 226, 2013, 367–385.
[6] T. Ertay And D. Ruan, “Data envelopment analysis based decision
model for optimal operator allocation in CMS,” European Journal
of Operational Research, vol. 164, 2005, pp. 800–810.
[7] J. Xu, X. Xu and S.Q. Xie, “Recent developments in Dual
Resource Constrained (DRC) system research,” European Journal
of Operational Research, vol. 215, 2011, 309–318.
[8] O. Battaia and A. Dolgui, “A taxonomy of line balancing problems
and their solution approaches,” International Journal of Production
Economics, http://dx.doi.org/10.1016/j.ijpe.2012.10.020, 2012.
[9] K. Nakade and R. Nishiwaki, “Optimal allocation of heterogeneous
workers in a U- shaped production line,” Journal of Computers
and Industrial Engineering, vol. 54, 2008, 432–440.
[10] J.T. Felan and T.D. Fry, “Multi-level heterogeneous worker
flexibility in a dual resource constrained (DRC) job-shop,”
International Journal of Production Research, vol. 39, 2001, pp.
3041–3059.
[11] Q. Wang, S. Lassalle, A.R. Mileham and G.W. Owen, “Analysis of
a linear walking worker line using a combination of computer
simulation and mathematical modeling approaches,” Journal of
Manufacturing Systems, vol. 28, 2009, pp. 64–70.
[12] L.I. Sennott, M.P. Van Oyen and S.M.R. Iravani, “Optimal
dynamic assignment of a flexible worker on an open production
line with specialists,” European Journal of Operational Research,
vol. 170, 2006, pp. 541–566.
[13] V.I. Cesani and H.J. Steudel, “A study of labor assignment
flexibility in cellular manufacturing systems”. Computers &
Industrial Engineering, vol. 48, 2005, pp. 571–591.
[14] P.M. Bobrowski and P.S. Park, “An evaluation of labor assignment
rules when workers are not perfectly interchangeable,” Journal of
Operations Management, vol. 11, 1993, pp. 257–268.
[15] M. J. Anzanello and F. S. Fogliatto, “Learning curve modeling of
work assignment in mass customized assembly lines,” International
journal of production research, vol. 45, 2007, pp. 2919–2938.
[16] K. Thongsanit, R. Boondisakulchok and W. Tharmmaphornphilas,
“Heuristic for Task-Worker Assignment with Varying Learning
Slopes,” Engineering Journal, vol. 14, 2010.
[17] S.M. Shafer, D.A. Nembhard and M.V. Uzumeri, “The Effects of
Worker Learning, Forgetting, and Heterogeneity on Assembly Line
Productivity,” Management Science, vol. 47, 2001, pp. 1639–1659.
[18] J.O. McClain, K.L. Schultz and L.J. Thomas, “Management of
worksharing systems,” Manufacturing & Service Operations
Management, vol. 2, 2000, pp. 49–67.
[19] A. Scholl and C. Becker, “State-of-the-art exact and heuristic
solution procedures for simple assembly line balancing,” European
Journal of Operational Research, vol. 168, 2006, pp. 666–693.
[20] C. Miralles, J.P. Garcia-Sabater , C. Andres and M. Cardos,
“Advantages of assembly lines in sheltered work centres for
disabled. A case study,” International Journal of Production
Economics, vol. 110, 2007, pp. 187–197.
[21] C. Miralles, J.P.G. Sabater , C. Andres and M. Cardos. “Branch
and bound procedures for solving the Assembly Line Worker
Assignment and Balancing Problem: Application to Sheltered
Work centers for Disabled,” Discrete Applied Mathematics, vol.
156, 2008, pp. 352–367.
[22] H.S. Ahn, I. Duenyas and M.E. Lewis, “The optimal control of a
two-stage tandem queueing system with flexible servers,”
Probability in the Engineering and Informational Sciences, vol. 16,
2002, pp. 453–469.
[23] R.G. Askin, and J. Chen, “Dynamic task assignment for throughput
maximization with worksharing,” European Journal of Operational
Research, vol. 168, 2006, pp. 853–869.
[24] J. Chen and R.G. Askin, “Throughput maximization in serial
production lines with worksharing,” International Journal of
Production Economics vol. 99, 2006, pp. 88–101.
[25] J.J. Bartholdi, L.A. Bunimovich and D.D. Eisenstein, “Dynamics
of 2- and 3-worker 'bucket brigade' production lines,” Operations
Research, vol. 47, 1999, pp. 488–491.
[26] J.J. Bartholdi, D.D. Eisenstein and R.D. Foley, “Performance of
bucket brigades when work is stochastic,” Operations Research,
vol. 49, 2001, pp. 710–719.
[27] A.I. Bratcu, and A. Dolgui, “A survey of the self-balancing
production lines (‘‘bucket brigades’’),” Journal of Intelligent
Manufacturing, vol. 16, 2005, pp. 139–158.
[28] D. Sparlingand and J. Miltenburg, “The mixed-model U-line
balancing problem,” International Journal of Production Research,
vol. 36, 1998, pp. 485–501.
[29] K. Nakade and K. Ohno, “An optimal worker allocation problem
for a U-shaped production line,” International Journal of
Production Economics, vol. 60, 1999, 353–358.
[30] E. Zavadlav, J.O McClain and L.J. Thomas, “Self-buffering, self-
balancing, self-flushing production lines,” Management Science,
vol. 42, 1996, pp. 1151–1164.
[31] G. Celano, A. Costaa, S. Ficheraa and G. Perroneb, “Human factor
policy testing in the sequencing of manual mixed model assembly
lines”. Computers & Operations Research, vol. 31, 2004, pp. 39–
59.
[32] C. Caux and H. Pierreval, “Regroupement de machines et
d'opérateurs en cellules par des algorithmes évolutionnistes,” Proc.
International conference in insdutrial eng. (CIGI), Montréal,
October18-20, 1995, pp. 163-172.
[33] G.A. Suer and C. Dagli, “Intra-cell manpower transfers and cell
loading in labor-intensive manufacturing cells,” Computers &
Industrial Engineering, vol. 48, 2005, pp. 643–655.
[34] B.A. Norman, W. Tharmmaphornphilas, K. L. Needy, B. Bidanda
and R.C. Warner, “Worker Assignment in Cellular Manufacturing
Considering Technical and Human Skills,” International Journal of
Production Research, vol. 40, 2002, pp. 1479–1492.
[35] H. Min And D. Shin, “Simultaneous Formation of Machine and
Human Cells in Group Technology: A Multiple Objective
Approach,” International Journal of Production research, vol. 31,
1993, vol. 2307–2318.
[36] M.B. Aryanezhad, V. Deljoo and S.M.J. Mirzapour Al-e-hashem,
“Dynamic cell formation and the worker assignment problem: a
new model,” International Journal of Advanced Manufacturing
Technology, vol. 41, 2009, pp. 329–342.
[37] H.V. Kher, “Examination of worker assignment and dispatching
rules for managing vital customer priorities in dual resource
constrained job shop environments,” Computers and Operations
Research, vol. 27, 2000, pp. 525–537.
[38] O.U. Araz and L. Salum, “A multi-criteria adaptive control scheme
based on neural networks and fuzzy inference for DRC
manufacturing systems,” International Journal of Production
Research, vol. 48, 2008, 1–20.
[39] Q. Li, J. Gong, J. Tang and J. Song, “Simulation of the model of
workers assignment in cellular manufacturing based on the
multifunctional workers,” Control and Decision Conference,
Chinese, 2008, pp. 992–996.
[40] M.C.O. Moreira and A.M. Costa, “A minimalist yet efficient tabu
search for balancing assembly lines with disabled workers,” In:
Anais do XLI Simposio Brasileiro de Pesquisa Operacional, Porto
Seguro, 2009.
[41] E. Gokhan and G.A. Suer, “Stochastic manpower allocation and
cell loading in cellular manufacturing systems,” 41st International
Conference on Computers & Industrial Engineering, Los Angeles,
CA USA, 2011.
[42] F. Nikoofarid and A. Aalaei, “Production planning and worker
assignment in a dynamic virtual cellular manufacturing system,”
International Journal of Management Science and Engineering
Management, vol. 7, 2012, pp. 89–95.
[43] J. Slomp, B.V. Chowdary and N.C. Suresh, “Design of virtual
manufacturing cells: a mathematical programming approach,”
Robotics and Computer-Integrated Manufacturing, vol. 21, 2005,
pp. 273–288.
[44] G.A. Widyadana, “Multi objective model for balancing U-type
assembly line with permanent and temporary workers,” Jurnal
Teknik Industri, vol. 11, 2009, pp. 33–42.
[45] C. Blum and C. Miralles, “On solving the assembly line worker
assignment and balancing problem via beam search,” Computers &
Operations Research, vol. 38, 2011, pp. 328–339.
[46] B.L. Song, W.K Wong, J.T. Fan, and S.F Chan, “A recursive
operator allocation approach for assembly line-balancing
optimization problem with the consideration of operator
efficiency,” Computers & Industrial Engineering, vol. 51, 2006,
pp. 585–608.
[47] G.M. Campbell and M. Diaby, “Development and evaluation of an
assignment heuristic for allocation cross-trained workers,”
European Journal of Operational Research, vol. 138, 2002, pp. 9–
20.
[48] H. ElMaraghy, V. Patel and I. Ben Abdallah, “Genetic algorithm
based approach for scheduling of dual-resource constrained
manufacturing systems,” CIRP Annals Manufacturing Technology,
vol. 48, 1999, pp. 369–372.
[49] H. ElMaraghy, V. Patel and I. Ben Abdallah, “Scheduling of
manufacturing systems under dual-resource constraints using
genetic algorithms,” Journal of Manufacturing Systems, vol. 19,
2000, pp. 186–198.
[50] A.A. Chaves, C. Miralles and L.A.N. Lorena, “Clustering search
approach for the assembly line worker assignment and balancing
problem,” Proceedings of the 37th International Conference on
Computers and Industrial Engineering, Alexandria, Egypt, 2007,
pp. 1469–1478.
[51] Z. Tao, T. Xiao and C. Hao, “Petri Net and GASA based approach
for dynamic JSP,” In: Proceedings of the IEEE International
Conference on Mechatronics and Automation, ICMA. Art. no.
4304196, 2007.
[52] D.P. Bischak, “Performance of a manufacturing module with
moving workers,” IIE Transactions, vol. 28, 1996, pp. 723– 733.
[53] L. Liu and Y. Li, “The fuzzy quadratic assignment problem with
penalty: New models and genetic algorithm,” Applied Mathematics
and Computation, vol .174, 2006, pp. 1229–1244.
[54] A. Azadeh, M. Anvari, B. Ziaei and K. Sadeghi, “An integrated
fuzzy DEA–fuzzy C-means simulation for optimization of operator
allocation in cellular manufacturing systems,” International journal
of advanced manufacturing technology, vol. 46, 2010, pp.361–375.
[55] A. Azadeh, B.P. Nokhandan, S.M. Asadzadeh and E. Fathi,
“Optimal allocation of operators in a cellular manufacturing system
by an integrated computer simulation–genetic algorithm approach,”
International Journal of Operational Research, vol. 10, 2011, pp.
333–360.
[56] A. Azadeh, S.N. Shirkouhi, L.H. Shirkouhi and A. Ansarinejad, “A
unique fuzzy multi-criteria decision making: computer simulation
approach for productive operators assignment in cellular
manufacturing systems with uncertainty and vagueness,”
International journal of advanced manufacturing technology, vol.
6, 20011, pp. 329–343.
[57] H. Pierreval, “Expert system for selecting priority rules in flexible
manufacturing systems,” Expert Systems with applications, vol. 5,
1999, pp. 51-57.
[58] H. Pierreval, Training a neural network by simulation for
dispatching problems, Proc. of the Third international conference
on CIM, Troy, New York, Mai 20-22, 1992, pp. 332–336.
[59] A. Longonia, R. Golinib and R. Cagliano, “The role of New Forms
of Work Organization in developing sustainability strategies in
operations,” International Journal of Production Economics, http
://dx.doi.org/10.1016/j.ijpe.2013.09.009.
... These types of problems have been widely studied in the literature about manufacturing systems and it is well known that the way how workers are assigned in the system greatly affects the system performances. Reviews of research works related to workers assignment in manufacturing systems are presented in (Ammar, Pierreval, & Elkosantini, 2013;Ernst, Jiang, Krishnamoorthy, & Sier, 2004;Van den Bergh, Beliën, De Bruecker, Demeulemeester, & De Boeck, 2013;Xu, Xu, & Xie, 2011). According to these literature reviews, there exist various definitions, assumptions, contexts and methods related to the workers assignment problems, which are known for their complexity and for their NP-Hardness (Mahdavi, Paydar, Kia, & Khonakdari, 2010;Sirovetnukul & Chutima, 2010). ...
... According to Ammar et al. (2013) and Xu et al. (2011), the majority of existing publications are concerned with problems where the set of jobs to be processed is defined in advance, with deterministic processing times. As a consequence, the task durations are known a priori, with enough certainty, and the completion dates of jobs can be computed. ...
... Given the lack of precision about the production plan to be performed, it is well known that usual optimization approaches can generally not be applied. As a consequence, several researchers, such as Ammar et al. (2013), Cesani and Steudel (2005) and Zavadlav, McClain, and Thomas (1996), have pointed out, from their literature analyses, that determining online, in real time, where the operators should be assigned can represent an alternative to react quickly to these changes. The assignment decisions are made, online, according to the current system state, with respect to a set of constraints (e.g. ...
Conference Paper
Facility layout problems (FLP) in manufacturing systems are widely addressed in the literature. FLP generally consist in finding the variables indicating the positions of the facilities, in order to favor (a) given objective(s), while respecting some constraints. They are known to have a significant impact on the performances of manufacturing systems. For instance, they can greatly impact the cost, productivity, works in process, and lead times of jobs in the system. Therefore, numerous studies have been conducted in the scientific literature to deal with this kind of problems and many approaches are proposed to solve it. In most approaches, researchers focus on classical production objectives of the manufacturing systems such as transportation and material handling costs. I, they mainly use technical measures to evaluate the performances associated with a given layout (e. g. distance between machines, adjacency score and material handling device utilization). However, the way, in which the facilities are arranged, does not only affect this type of measures: it can also impact human productivity, behaviors and even health, which can also affect the system efficiency. Unfortunately, it seems that human factors, such as workers’ fatigue, are insufficiently taken into account in FLP. In this study, we are particularly interested to the workers’ fatigue because it can lead to negative effects on workers (e.g. performances and musculo-skeletal troubles) and efficiency of production systems (e.g. deterioration of processing times and an increase in the flowtime). We are more particularly interested in taking into account the workers’ fatigue in the evaluation of layout designs. In this respect, we need to understand how a layout can contribute to the workers’ fatigue, in view of trying to avoid or to reduce its negative effects. For that, it is important to identify the fatigue factors related to FLP. Indeed, the factors that lead to the workers’ fatigue and can affect the design of layouts, particularly when workers are multi-skilled, are mainly related to the work arduousness such as noise, uncomfortable postures, vibrations and heavy loads. When workers are exposed to arduousness, task duration can also greatly impact the workers’ fatigue. In order to understand the consequences of how manufacturing facilities are arranged in terms of efficiency and to understand how fatigue emerges and evolves along time if a given layout is chosen, we need to be able to evaluate possible layout solutions in order to define an appropriate design. In this respect, simulation turns out to be an interesting approach. Since the evolution of fatigue is dynamic and continuous, static criterion may be not suited. We need to dynamically describe the evolution of the workers’ fatigue (accumulation and recovery) so as to analyze the evolution trends of fatigue. This implies the use of a combined simulation model (discrete and continuous simulation) and to be able to specify how fatigue evolves along time. In order to quantify the workers’ fatigue, we can use one of the few fatigue mathematical models, which are proposed in the ergonomic literature. In this study, we use the fatigue model (Eq (1) and (2)), which is widely used in the literature. Eq (1) represents the fatigue accumulation function and Eq(2) represents recovery, where 𝐹𝑖(𝑡) indicates the fatigue of the worker i at time t. Fi(t) = 1 ‒ e-λt (1) R(θ) = Fi(t) e-μt (2) The variables 𝜆 and µ indicate, respectively, the speed of accumulation and recovery. They depend on the fatigue factors, which are mentioned in the previous paragraph, and recovery factors (i.e. rest period and the physical characteristics of each worker for recovery (e.g. sex and age)). However, it is often difficult to have accurate information about how these factors can be considered so to determine 𝜆 and µ. Thus, fuzzy logic concepts could be used to deal with the problem of inaccuracy. We suggest representing the accumulation and recovery factors, in our simulation model, by linguistic variables and how they can impact the evolution of fatigue can be defined by fuzzy rules. For instance, IF noise is high and load is heavy THEN the accumulation speed is high. Regarding the evaluation of the layout efficiency, several criteria can be used such as the flowtime of jobs in the system and the work in process. In this study, we focus on the mean flowtime of jobs (MFT) because it is known as an interesting global performance measure and because operating times of tasks can increase due to the workers’ fatigue, so that they can take longer than expected. Let Tkj be the “theoretical” (expected) processing time of the task waiting the product k to be processed on machine j. In order to take into consideration the impact of fatigue on Tkj, we correct it as suggested in another previous work, such that T’kj(t) in (3) is the actual processing time and δ is a parameter used to tune the effect of fatigue. T’kj(t) = Tkj (1+λ δ (1+Fi(t))) (3) In order to illustrate our purpose and highlight the impact of the layout on the system performances and on the workers’ fatigue, we propose to compare two different layout designs of a job-shop system. To select the most appropriate design, we use a simulation model of a job-shop system and we are interested in choosing the design that reduces system efficiency while avoiding the excessive fatigue of the workers. Simulation results show the importance of taking fatigue into consideration to make better decisions about layouts so to improve system efficiency (MFT) and workers’ well-being at work
... These types of problems have been widely studied in the literature about manufacturing systems and it is well known that the way how workers are assigned in the system greatly affects the system performances. Reviews of research works related to workers assignment in manufacturing systems are presented in (Ammar, Pierreval, & Elkosantini, 2013;Ernst, Jiang, Krishnamoorthy, & Sier, 2004;Van den Bergh, Beliën, De Bruecker, Demeulemeester, & De Boeck, 2013;Xu, Xu, & Xie, 2011). According to these literature reviews, there exist various definitions, assumptions, contexts and methods related to the workers assignment problems, which are known for their complexity and for their NP-Hardness (Mahdavi, Paydar, Kia, & Khonakdari, 2010;Sirovetnukul & Chutima, 2010). ...
... According to Ammar et al. (2013) and Xu et al. (2011), the majority of existing publications are concerned with problems where the set of jobs to be processed is defined in advance, with deterministic processing times. As a consequence, the task durations are known a priori, with enough certainty, and the completion dates of jobs can be computed. ...
... Given the lack of precision about the production plan to be performed, it is well known that usual optimization approaches can generally not be applied. As a consequence, several researchers, such as Ammar et al. (2013), Cesani and Steudel (2005) and Zavadlav, McClain, and Thomas (1996), have pointed out, from their literature analyses, that determining online, in real time, where the operators should be assigned can represent an alternative to react quickly to these changes. The assignment decisions are made, online, according to the current system state, with respect to a set of constraints (e.g. ...
Conference Paper
It is known that the workers’ fatigue can greatly affect both industrial performances and the workers’ wellbeing at work. Consequently, when designing a manufacturing system, managers are interested in facility layouts that favor performance objectives, while avoiding excessive fatigue. Unfortunately, most existing studies related to layout design focus on technical aspects of the considered system (e.g. flow costs, distance between machines, etc.) so that human factors, in particular fatigue, are insufficiently taken into consideration. Therefore, we are interested in how the workers’ fatigue can be taken into account when evaluating possible layout designs. We analyze the factors that induce fatigue, which are mainly concerned with the work arduousness, and depend on the layout. We explain how they can be considered in order to compare possible solutions of a layout problem. In such a context, emphasis is put on the role of simulation. We illustrate our purpose and highlight the importance of taking fatigue into consideration through a comparison, using simulation, of two different layouts of a jobshop system. The comparison is based both on the mean flowtime of jobs and how the workers’ fatigue evolves over time.
... These types of problems have been widely studied in the literature about manufacturing systems and it is well known that the way how workers are assigned in the system greatly affects the system performances. Reviews of research works related to workers assignment in manufacturing systems are presented in (Ammar, Pierreval, & Elkosantini, 2013;Ernst, Jiang, Krishnamoorthy, & Sier, 2004;Van den Bergh, Beliën, De Bruecker, Demeulemeester, & De Boeck, 2013;Xu, Xu, & Xie, 2011). According to these literature reviews, there exist various definitions, assumptions, contexts and methods related to the workers assignment problems, which are known for their complexity and for their NP-Hardness (Mahdavi, Paydar, Kia, & Khonakdari, 2010;Sirovetnukul & Chutima, 2010). ...
... According to Ammar et al. (2013) and Xu et al. (2011), the majority of existing publications are concerned with problems where the set of jobs to be processed is defined in advance, with deterministic processing times. As a consequence, the task durations are known a priori, with enough certainty, and the completion dates of jobs can be computed. ...
... Given the lack of precision about the production plan to be performed, it is well known that usual optimization approaches can generally not be applied. As a consequence, several researchers, such as Ammar et al. (2013), Cesani and Steudel (2005) and Zavadlav, McClain, and Thomas (1996), have pointed out, from their literature analyses, that determining online, in real time, where the operators should be assigned can represent an alternative to react quickly to these changes. The assignment decisions are made, online, according to the current system state, with respect to a set of constraints (e.g. ...
Article
Manufacturing systems are often characterized by a stochastic and uncertain behavior in which frequent changes and unpredictable events may occur over time. Moreover, the customers’ demands can sometime evolve drastically along time. In order to cope with such changes in the manufacturing system state, and to optimize given performance criteria, the assignment of multi-skilled workers to the machines in the system can be decided online, in a dynamic manner, whenever workers become available and need to be assigned. Indeed, the starting and completion times of jobs in such systems cannot be predicted, so that static optimization approaches turn out not to be relevant. Several studies, in the ergonomics literature, have outlined that the operators' performances often decline because of their fatigue in work. In particular, in manufacturing contexts, fatigue can increase the processing times of jobs. Several online heuristic have been published, but to the best of our knowledge, they do not cope with this consequence of fatigue. We propose to solve this dynamic multi-skilled workers assignment problem using a new methodology, which aims to provide an adaptable dynamic assignment heuristic, which is used online. Our approach takes the impact of fatigue into consideration, in order to minimize the mean flowtime of jobs in the system. We suggest computing more realistic task durations, in accordance with the worker's fatigue. The heuristic uses a multi-criteria analysis, in order to find a compromise that favors short processing times and avoids congestions. The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method is used to select the machine where to assign the worker. Since in our case no expertise is available, an offline adaptation process, based on simulation optimization, is used to identify the weights needed by TOPSIS, so as to better fit with the system specificities. A Job-Shop systems are simulated to illustrate the proposed approach. The performance of the suggested heuristic is assessed and compared to two other workers assignment rules, which are widely used in the scientific literature because of their efficiency on the mean flowtime: SPT and LNQ. The comparisons are made under different conditions (staffing level, operators’ flexibility). A sensitivity analysis is also performed to analyze the impact of the way how fatigue affects the task duration. Our experimental results show that our heuristic provides better results in every case studied. Several important research directions are finally pointed out.
... Many manufacturing systems are characterized by a predefined arrival time of orders and constant operating times. In such cases, static assignment is used, so that workers could be assigned, for a single period, to the different machines in the system [3]. According to Ammar et al. [3] and Xu et al. [4], which have analysed the existing literature on the workforce assignment problems, classical optimization methods (i.e. ...
... In such cases, static assignment is used, so that workers could be assigned, for a single period, to the different machines in the system [3]. According to Ammar et al. [3] and Xu et al. [4], which have analysed the existing literature on the workforce assignment problems, classical optimization methods (i.e. exact, heuristics and metaheuristics approaches) have been extensively used to solve the static assignment of workers. ...
... In order to react quickly to these changes, workers can be assigned dynamically, online and according to the current system state. For such types of workers assignment problems, Ammar et al. [3] point out that the assignment rules When and Where are the most used in the literature. The When-rule specifies when the worker is eligible for transfer and the Where-rule specifies the machine to which the worker is going to move once he/she is released [7]. ...
Conference Paper
In order to cope with the frequent unpredictable changes that may occur in manufacturing systems, and to optimize given performance criteria, the assignment of workers can be decided online in a dynamic manner, whenever the worker is released. Several studies, in the ergonomics literature, have shown that individuals' performances decrease because of their fatigue in work. In manufacturing context, the workers' fatigue impacts the task durations. Therefore, we propose to solve the online workers assignment problem through a heuristic, which takes this workers' fatigue into consideration, so as to minimize the mean flowtime of jobs. This approach suggests computing more realistic task duration in accordance with the worker's fatigue and it uses multi-criteria analysis in order to find a compromise to favor short durations and to avoid congestions. The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method is used to select where to assign the worker. A learning process through simulation optimization is used to adapt the weights, in TOPSIS, to better fit with the system characteristics. The approach is illustrated with a simulated Job-Shop system. Experimental results comparing our approach with the rule Shortest Processing Time (SPT), which is known as efficient on the mean flowtime, show the effectiveness of the heuristic.
... A recent survey of existing research about workers assignment shows that most works address static problems where tasks or machines are known and fixed for a given time period. Unfortunately, in many real situations, the frequent changes in the production, the unexpected disturbances and the arrival of urgent orders can make static approaches irrelevant [1]. For example, when the system has a stochastic behavior, the bottleneck can move from one work center to another, which implies the change of workers assignment. ...
... According to [1], most articles dealing with workers assignment address static cases where the production system features are defined by a fixed number of workers, each machine capacity being known and constant, workers are generally assigned for a single time period with deterministic and constant product mix and demand so that unexpected events are not considered. As a consequence, such problems are generally solved using classical optimization methods (either exact, such as branch and bound, or inexact such as heuristic methods). ...
Conference Paper
Full-text available
Assigning workers to machines dynamically, so as to achieve production objectives, is known as a difficult problem which must be addressed in several manufacturing systems. In fact, production managers must determine when and where to assign workers, which is a very difficult task because of the complexity induced by the dynamic changes and the stochastic behavior of these systems. The current scientific literature is mainly oriented towards static assignments problems (only few dynamic heuristics are published). We propose a new approach whose aim is to help the managers in improving their knowledge about how to assign workers in real time. It is based on Serious Games built using Visual Interactive Simulation (discrete event). Our experiments show that an appropriate training using the proposed game can improve the users' ability to make good decisions about dynamic workers assignment and that human decisions can become more efficient than existing workers assignment rules found in the literature.
... WS is one of the major problems that manufacturing companies have to face, in which workers can be assigned to one or several tasks within a certain period of time (Ammar & Elkosantin, 2013). The WS problem occurs when the number of available workers is less than the number of operations, as in dual resources constrained (DRC) systems where workers can be relocated across various workstations (Xu et al., 2015). ...
Article
Full-text available
This work is focused on workforce scheduling for assembly lines with the additional constraint of workforce distancing. The aim is to warrant the necessary safety and health requirements due to COVID-19. The research stems within an industrial case in which a methodology has been developed with the objectives of i) developing a constraint optimization model considering the social distancing of workers as part of the workforce scheduling requirements and ii) investigating how the workforce distancing can affect certain production performances. Through an empirical investigation the impact of distancing on workforce allocation is appraised in terms of daily production capacity. Then, different distancing thresholds are assessed to seek the optimal balance among production performances and safety requirements. The research resulted in a tool able to adapt the scheduling sequence to those health/safetysituations where the production manager needs to minimize losses in terms of production capacity, warranting the safest working conditions.
... Most of the other studies were done in the context of production scheduling problems. Several articles provided the states of the art on the existing advances in workforce planning research [Ammar et al., 2013, De Bruecker et al., 2015]. ...
Thesis
Mass customization and frequent market fluctuations push industrial companies to employ flexible and reconfigurable multi/mixed-model assembly lines instead of dedicated ones. This thesis focuses on this problem. It concentrates mainly on mixed-model assembly line design and balancing problems. The questions concerning the efficiency of such lines, the importance of optimal task assignment and use of walking workers are asked and studied. To increase the flexibility of the line, we account for different types of task assignments: fixed, model-dependent,and dynamic. We aim to design a line that can handle various entering product models. We use combinatorial optimization methods, and,in particular, robust optimization approaches. We present an extensive literature review on line balancing, workforce planning, and workforce reconfiguration strategies in different production systems. The first problem addresses a configuration selection problem between a single multi-model line and multiple dedicated lines. The second problem consists in designing and balancing a mixed-model assembly line with walking workers. We propose fixed and model-dependent task assignments for a given set of product mixes. The goal is to minimize the total cost of workers and equipment for the worst case. The third problem extends the second one. It considers the dynamic task assignment. In the last problem, we extend the third problem for the case where the sequence of products unfolds takt by takt. In this context, we minimize both the expected total cost and the worst-case cost. In order to solve the considered problems, we develop several exact methods and heuristics: mixed-integer linear programming models, greedy algorithm, local search,matheuristic and fixed-and-optimize heuristics among others. We also apply a Markov Decision Process to the proposed line balancing problem in the last chapter. It is the first study applying this method to a line balancing problem. Computational experiments evaluate the performance of the proposed approaches in terms of solution quality and time consumption. We draw managerial insights in each chapter. Our results show the superiority of the dynamic task assignment compared to model-dependent and fixed ones in different production situations.
... • Human factors and ergonomics need to be considered at configuration design stage in order to take into account operator skills, preferences and motivation in 2017 International Conference on Advanced Systems and Electric Technologies (IC_ASET) the assignment of operators to tasks within configurations [32]; ...
Conference Paper
Full-text available
In reconfigurable manufacturing systems (RMS), the evaluation of configurations is usually based exclusively on technical performance indicators for an efficient and effective manufacturing. Ergonomics and human factors related aspects are rarely considered in the evaluation of reconfiguration opportunities, which leads to less realistic and pragmatic reconfiguration decisions. This article suggests both technical and ergonomics indicators to achieve a more realistic evaluation of reconfiguration decisions. A case study is introduced, and a TOPSIS based method is used to achieve multi-criteria evaluation and selection of an alternative configuration when reconfiguring an RMS. We particularly compare reconfiguration decisions with and without ergonomics indicators, and therefore establish the worthiness of considering both aspects simultaneously.
... En cohérence avec (Ammar et al. 2013), nous n'avons pas identifié de travaux utilisant des méthodes multicritères de choix, ni d'optimisation via simulation pour définir des méthodes d'affectation. Ces deux approches sont à la base de l'heuristique présentée dans le paragraphe suivant. ...
Article
In this paper, we introduce a mixed-integer linear program for a shift scheduling problem in a German potash mine. In particular, we consider a short-term (work shift) production scheduling problem, where drill-and-blast mining operations have to be assigned to machines and workers simultaneously. Since we deal with several sequence-dependent setup, changeover, and removal times, TSP-variables are used in the mathematical program to determine the processing-sequence of the operations on each worker and each machine, respectively. In addition, several mining-specific requirements are taken into account to obtain a solution that can be put into practice. Computational experiments are conducted on problem instances of realistic size derived from real-world data. The results show that our new mixed-integer linear formulation outperforms both existing solution procedures for the problem at hand.
Conference Paper
Full-text available
In this paper, a stochastic manpower allocation and cell loading problem is studied, where operation times and customer demand are probabilistic. A two-phase hierarchical methodology is proposed to find the optimal manpower assignment and cell loads simultaneously with respect to maximum allowable risk levels. Two non-linear stochastic mathematical models are developed to deal with manpower allocation and cell loading phases. In both models, processing times and demand are normally distributed. Firstly, alternative configurations are generated. Secondly, stochastic manpower allocation is performed and products are loaded to cells. Statistical analysis and Monte-Carlo simulation are employed to link the production rates obtained from the first phase to the capacity requirements that are used in the second phase. The methodology is illustrated with an example problem drawn from a real manufacturing company. According to the results, as risk level increases, lower number of manpower is assigned to cells. Expected utilizations of cells are decreased as the variance of demand and processing times increase. The approach allows decision maker to perform manpower allocation with respect to the desired risk level.
Article
Full-text available
We describe all possible asymptotic behavior of bucket brigade produc-tion lines with two or three workers, each c haracterized by a constant w ork velocity. The results suggest wariness in interpreting simulation results. They also suggest a strategy for partitioning a workforce into eeective teams to staa the lines.
Article
Full-text available
This paper aims to study whether Human Resource Management and the organizational practices related to New Forms of Work Organization (NFWO) (e.g., teamwork, training, and employee involvement) should be implemented to attain higher environmental and social sustainability performance. This potential connection is analyzed using the International Manufacturing Strategy Survey 2009 database containing data on the assembly industry. Several hierarchical regression models are used to study the direct impact of the considered practices on sustainability performance and their interactions with sustainability action programs. The results show that some of the practices related to NFWO are linked to sustainability performance. In particular, training has a direct positive effect on environmental and social sustainability performance and creates a positive interaction between social sustainability action programs and performance. Additionally, employee involvement and incentives have a direct positive impact on social sustainability performance. Finally, teamwork is a relevant practice for the successful implementation of environmental sustainability action programs. This paper contributes by empirically extending the knowledge on the role of organizational practices and the understanding of environmental and social sustainability achievement at the operational level. Moreover, we highlight the complexity of these relationships within different sustainability dimensions, showing the need for more qualitative studies about this topic.
Article
Full-text available
Line balancing belongs to a class of intensively studied combinatorial optimization problems known to be NP-hard in general. For several decades, the core problem originally introduced for manual assembly has been extended to suit robotic, machining and disassembly contexts. However, despite various industrial environments and line configurations, often quite similar or even identical mathematical models have been developed. The objective of this survey is to analyze recent research on balancing flow lines within many different industrial contexts in order to classify and compare the means for input data modelling, constraints and objective functions used. This survey covers about 300 studies on line balancing problems. Particular attention is paid to recent publications that have appeared in 2007–2012 to focus on new advances in the state-of-the-art.
Article
We model a U-shaped manufacturing module that operates with fewer workers than machines. Workers move in order around the layout. A definition of the module's rules of movement are presented. The throughput and expected average throughput time of this module, with and without buffers, are compared with those of a serial line with one stationary worker per machine. Simulation results indicate that movingworker modules can provide flexible capacity without the use of buffers. In unbuffered systems with high processing time variation the moving-worker module can have higher throughput than the stationary-worker system. Recommendations for the design of these modules are given.
Article
Many manufacturers are switching their production lines from single product or batch production to mixed-model production, often as a consequence of implementing just-in-time (JIT) principles into their operations. In mixed-model production, different products or models are produced on the same line with the models interspersed throughout a production sequence. This helps manufacturers provide their customers with a variety of products in a timely and cost-effective manner. The mixed-model U-line balancing (MMULB) problem assigns the tasks required to produce all models to a minimum number of stations on a U-shaped line. U-lines are widely used and well-suited to mixed-model production. A model of the MMULB problem is developed in this paper. The problem is NP-hard. An approximate solution algorithm is presented, and an illustrative example is worked. Areas where more research is needed are identified.
Article
Most traditional cell formation problem (CFP) procedures disregard any changes in demand and part mix variation. However, in today’s business environment, the life cycle of goods is short, and demand volumes and product mix can frequently vary. In this article, a mathematical model for production planning in a dynamic virtual environment is proposed which includes an extensive coverage of the important manufacturing features, including a consideration of multi-period planning horizons with demand and part mix variation, machine capacity, the main constraints are demand satisfaction in all over period, machine availability, machine time-capacity, and availability of workers. VCMS combines the robustness and flexibility of process orientation with the advantages of product orientation, which is expected to appreciate throughput time performance. To validate and verify the proposed model, computational results are presented by solving a number of numerical examples, with a linearized formulation.
Article
This paper presents a review of the literature on personnel scheduling problems. Firstly, we discuss the classification methods in former review papers. Secondly, we evaluate the literature in the many fields that are related to either the problem setting or the technical features. Each perspective is presented as a table in which the classification is displayed. This method facilitates the identification of manuscripts related to the reader’s specific interests. Throughout the literature review, we identify trends in research on personnel staffing and scheduling, and we indicate which areas should be subject to future research.