Filetype PDF | Posted on 06 Feb 2023 | 3 years ago
Authentication
digital resources
314x Filetype PDF File size 0.34 MB Source: iaeme.com
International Journal of Mechanical Engineering and Technology (IJMET)
Volume 9, Issue 10, October 2018, pp. 1021–1032, Article ID: IJMET_09_10_105
Available online at http://iaeme.com/Home/issue/IJMET?Volume=9&Issue=10
ISSN Print: 0976-6340 and ISSN Online: 0976-6359
© IAEME Publication Scopus Indexed
INTEGRATED MODEL FOR MACHINE
SCHEDULING AND INVENTORY
MANAGEMENT UNDER FINITE CAPACITY
SETTINGS
V Mahesh
Department of Mechanical Engineering,
S R Engineering College, Warangal, Telangana, India.
ABSTRACT
The increasing importance of engineer-to-order manufacturing, inventory
reduction, and resource optimization have placed a significant emphasis on the growing
role of scheduling in the current global business environment. Often there has been a
problem of applying the mathematically rich theoretical concepts of scheduling theory
to solve the real-time industrial problems - in particular, the problem of job-shop
scheduling. In the manufacturing domain, many scheduling models and approaches
have been developed and adopted earlier. In most of the prior extant literature,
materials requirement and capacity requirement planning problems are solved
independently without considering the scheduling requirements. In such cases,
rescheduling needs to be carried due to the problems that arise from resource conflicts
and change in the job priorities. Therefore, there is a necessity to deal with scheduling,
material, and capacity planning in an integrated way. In this paper, the author
successfully demonstrated integration of scheduling with material requirement
planning (MRP) and capacity requirements planning (CRP), to generate a near to
optimal production schedule at low cost considering the practical difficulties in a real
job shop environment.
Keywords: Scheduling, Material requirement planning, Capacity requirement
planning.
Cite this Article Md. V Mahesh, Integrated Model for Machine Scheduling and
Inventory Management under Finite Capacity Settings, International Journal of
Mechanical Engineering and Technology, 10(10), 2018, pp. (1021)-(1032).
http://www.iaeme.com/IJMET/issues.asp?JTypeIJMETT&VType=6&IType=7
http://www.iaeme.com/IJMETT/index.asp 1021 editor@iaeme.com
V Mahesh
1. INTRODUCTION
Planning, scheduling, and control are the three iterative processes that should occur
continuously in production processing. These three processes influence and impact each other.
The material requirement planning (MRP), capacity requirements planning (CRP) and schedule
control, are the typical planning and control systems that arise from the above three production
activities and they should be integrated [1]. Irrespective of the kind of change occurring in
material requirement or capacity availability, it invariably ends in the adjustment of production
operations. If it is not agreeable, a rescheduling of the whole production process must be
undertaken to ensure the uninterrupted flow of operations. Since the reasons for rescheduling
can stem from a myriad of sources, it is necessary to consider all the related factors arising from
the various production activities for an optimum revision. The proposal to integrate related
production activities hinges upon the later point.
The integration that connects related production activities greatly facilitates large-scale
scheduling and significantly simplifies complex schedule in a bid more readily to coordinate an
entire manufacturing environment. Many manufacturing industries face the problem of
scheduling job shops and still finding methods for improvement. Generating a good schedule
for a multi-product manufacturing industry in a reasonable time remains a challenging problem,
because of the NP-hard nature of job scheduling and its inherent computational complexity.
Unfortunately, MRP-I does not consider capacity at all, while the initial promises of the
more extensive manufacturing resource planning (MRP II) system are not fulfilled. The rough-
cut capacity planning (RCCP) module of MRP II concerns only the long-term capacity
availability on a high aggregation level while capacity requirements planning (CRP) performs
just a check on the amount of capacity needed. In case of a mismatch between available and
required capacity, it is left to the planner to adjust the MPS. The temporary alteration of lot
sizes offers another possibility, but the difficulties in doing this in the formal system force a
planner to rely on informal procedures, thereby undermining the data accuracy and optimality
of the MRP system.
A capacity-oriented MRP procedure generates feasible plans of orders without requiring
lead-times as input and without relevant computational burden [2]. There are two types of
integrated models in the literature. The first way is to present models where one function is
basically considered while taking into account the other. These are the interrelated models. The
second way is to model two or more elements of the production system simultaneously. These
are the integrated models [3]. The findings of the successful implementation of computational
simulators for finite capacity scheduling like Issues regarding human, organisational and
technological aspects were highlighted throughout the implementation process and proved
essential for the solution to be effective [4]. James et al. proposed a heuristic capacity planning
algorithm which allocates orders to resources, determines appropriate order release time to the
factory, and estimates the expected loading of all machines [5]. Wuttipornpun has developed
an algorithm of finite capacity material requirement planning (FCMRP) system for a multistage
assembly flow shop. The study considers only lot-sizing policy, and the effect of different lot-
sizing policies has not been studied [6]. Satish et al. discussed the complete routing flexibility
with machine change and alternate machining process in a flexible manufacturing environment
and enhance the system performance [7, 8]. Mahesh et al proposed a computationally effective
powers-of-two heuristic for solving a job shop scheduling problem. The authors prove that the
makespan of the schedule obtained through powers-of-two release dates lies within 6% of the
optimal value [9]. A common and often commented upon the form of fixation is a premature
commitment to a particular problem solution. Consequently, the designer or planner stops
pursuing the search for alternative solutions. This premature commitment thus results in fewer
solutions [10].
Divergent thinking (DT) is one of the design skills which helps in the generation
http://iaeme.com/Home/journal/IJMET 1022 editor@iaeme.com
Integrated Model for Machine Scheduling and Inventory Management under Finite Capacity
Settings
of alternative design solutions for any given design task and is very much essential for
addressing a design problem [11]. A similar concept can be used by a planner and apply these
DT skills even for solving scheduling problems.
In most of the manufacturing industries material requirements planning, scheduling and
capacity utilization is done independently. If the information is thoroughly shared by integrating
three activities of the planning department, one can optimize time, resources and money as well.
Thus, integration of MRP, CRP with scheduling is an exciting area of research. It is to be
addressed thoroughly.
2. AN INDUSTRIAL ILLUSTRATION OF THE PROBLEM
A brief description of the problem, the terminology that is used in the database, complexity,
and magnitude of the scheduling problem is discussed in detail in the following section.
Each customer order is identified by a unique work order number (W ). In turn, each W
N N
consists of several (sub) assemblies, and they were identified by a unique number termed as the
product group assembly (PGA). Further, each PGA consists of individual parts identified by
their part number (P ). The sequence of operations of each part is identified by operation
N
number (O ). It should be noted that O would be a mapping of the part on to the machine with
N N
a time component. On an average, each part undergoes 25 to 125 operations. Further, the
recirculations and preemptions are quite common apart from the machine availability, tooling
breakdown etc. The industry typically handles more than one workplace (machine) in a work
center - that is, a group of similar machines, can be of varying capacity. Work centers are
identified by a number - W and the underlying workplaces are identified as W .
C P
A steam turbine manufacturing industry is considered to be a complex scheduling problem.
A typical steam turbine consists of approximately 800 components, which may be classified
into, major, sub-major, and minor components. The industry maintains the history of process
planning of all the work orders that were processed on the shop floor. A new work order details
are appended into a large-scale manufacturing enterprise system. The process plan of the work
order is developed by retrieving the information from an equivalent work order from the
enterprise. The complexity of generating a typical process plan would vary greatly. For work
orders, which are exactly similar in terms of routing and processing requirements, this task will
involve a copy, but if the new work order has a different sequence of operations then a non-
trivial engineering effort is required and hence increases process planning time.
3. DATA REPRESENTATION
A logical representation of data is very important to perform functions such as material
requirements planning (MRP), capacity requirements planning (CRP), operation scheduling
and shop floor control [12]. Hence an effective production planning and control system requires
combining the bill of material (BOM) and routing data to reflect the material flow through the
production process. An integrated BOM and routing data model allows the flexibility in
handling relationships between materials and operations to suit specific needs. It can also be
used as a standard data resource for creating production jobs [13].
Product represented by a BOM can be used for describing an end product to state raw
materials and intermediate parts or subassemblies required for making the product. Production
data is concerned with how a product is produced, i.e., it specifies the operation sequence and
the machines required for each operation. Similar to describing a product structure using BOM,
a bill-of-operations (BOO) can be constructed to represent the production structure of a given
product.
http://iaeme.com/Home/journal/IJMET 1023 editor@iaeme.com
V Mahesh
Traditionally BOMs and BOOs are treated as separate data files by most computer-based
production systems. The BOM being primarily responsible for MRP and inventory
management and the BOO being responsible for capacity requirements planning (CRP) and
Production Control. However, a job is a statement of product, which require both BOM and
BOO data. Effective control of a production job cannot be fulfilled without the integration of
planning and control functions [14]. Many authors have demonstrated the merits of integrating
the BOM and the BOO in production planning and control [15,16,17].
A bill of manufacture (BOMfr) is developed by combining the BOM structure and BOO
structure. The BOMfr specifies the sequence of production operations, as well as the materials
and resources required at each operation for making a product. In this way, the unification of
BOM and BOO can be achieved in a BOMfr structure.
The industry produces various types of turbines; the processing sequences of major parts of
these turbines are mostly the same. The user industry maintains the history of all the work orders
that were processed on the shop floor in the form of BOMfr. Hence, whenever a new work
order enters into the system, the manufacturing requirements (operation routing, material
requirements, etc.,) of it are copied from the BOMfr of an equivalent work order available in
the database. If the new work order has a different requirement then it takes non-trivial
engineering effort and hence increased time in process planning. Additional job attributes such
as a number of pieces, due date, and job precedence constraints are also created during the
copying process. This production job data form a basis for detailed operations scheduling and
shop floor control. This provides a planning standard for making standard or recurring products.
Table 1 given below represents few records of the BOMfr.
Table 1 BOMfr data illustrating the requirements of work orders 1 & 2 (sample data)
Processing
Material Weight per
W PGA P O W W
N N N C P
Required unit (Kg)
time (min)
Cast carbon
1 30101 1001 1 1032 9863 480 1000
steel
1 30125 25001 5 3116 9412 480 Grey Cast Iron 1055
1 30125 25058 2 3116 4828 720 Grey Cast Iron 1335
Alloy Steel
1 30127 27001 1 3116 9991 330 1345
Forgings
Alloy Steel
1 30201 1001 1 3116 9863 480 2019
Forgings
Alloy Steel
1 30209 9001 4 3112 9421 220 926
Castings
1 30301 1001 1 3112 2852 340 Nickel Steel 1000
1 30515 15001 3 3112 9863 480 Bronze Lining 415
Alloy Steel
1 30528 28001 6 3117 8577 500 846
Castings
Cast carbon
2 30101 1001 1 3117 9863 480 1000
steel
2 30125 25001 5 3116 9412 690 Grey Cast Iron 1055
2 30125 25058 2 3116 4828 1560 Grey Cast Iron 1335
Alloy Steel
2 30127 27001 1 3116 9991 750 1345
Forgings
Alloy Steel
2 30201 1001 1 3116 4828 1280 2019
Forgings
http://iaeme.com/Home/journal/IJMET 1024 editor@iaeme.com
no reviews yet
Please Login to review.
