Vous êtes sur la page 1sur 7

COMPARISON OF MAIN

ALGORITHM AND DELAYED


PRODUCT DIFFERENTIATION IN
LAYOUT DESIGNS

SUBMITTED BY
LIJO JOHN
M100298ME
IEM, MED
1. Introduction

Group Technology (GT) has been applied for generating the machine groups and
corresponding part families for the Cellular manufacturing (CM). We very well know that the
there is a close relationship between the product to be manufactured and the layout for its
manufacturing. In present day volatile market were the demand variation is a very big
challenge faced by the manufacturing companies. To cope up with this quantitative challenge
there are different layouts being suggested. This is a continuous effort to come out with
more and more better layouts which are technologically feasible and economically viable.
But one cannot forget amount of qualitative variation demanded by the customer. In such
case it becomes more difficult to design a layout which will give larger product variation.

Here, in this report I would like to compare two layouts which try to include the quantitative
and qualitative product variation in the cellular layout. The MAIN algorithm tries to optimize
the layout fro quantitative variation in demand and by using Delayed Product Differentiation
how we can optimize the layout for qualitative variation.

2. MAIN Algorithm

Here the author, Chan et.al, tries to investigate a Dynamic Machine Cellular Layout (DMCL)
in an environment with multiple periods planning horizon. Initially the machine layout will
be generated on the fixed quantitative demand profile and its being referred to as Static
Machine Cellular Layout (SMCL). Each SMCL is best suited for the planning period but in
actual case the demand seldom remains constant, and so does the product mix. DMCL is an
attempt to address the demand changes with unchanged product operating sequence in the
cell.

The author has addressed this problem of evaluating the most optimum layout for the given
planning horizon given the variation in the demand profile and thereby changing product mix
by considering the factors like machine rearrangement cost and the part travelling cost. In
addition to these factors a part travelling score was also formulated in the research to
compare the various SMCL’s which from the starting solution for the DMCL design.

The main objective is to reduce the total part travelling cost by taking into consideration the
closeness of the machine, the part handling efficiency, the part travelling distance, the
materials flow frequency, and the machine rearrangement cost in the dynamic environment.
The MAIN algorithm can be explained in brief like this. Initially the SMCL is formed for
different demand patterns in the different periods of the multiple period planning horizon
based on several factors like operation sequences, part flow frequencies, and the customer
demand of the part family in order to determine the required closeness between the machines
in the layout. The part handling score will be calculated for each SMCL layout based on the
related pair wise machines, each of which incorporates the information such as the distance,
the frequency of the part flow and the part handling efficiency. This score is being used to
compare different layouts and forms the basis for the development of the DMCL which
solely tries to reduce the layout cost which can be achieved by a balance between the total
machine rearrangement cost and the total part handling cost.

The problem considered here assumed 9 machines arranged in a regular 3x3 grid, each
machine requiring same floor space. For the model the part travelling distance is being
considered from the center of one cell to the center of another cell. In MAIN algorithm,
DMCL is being divided into two, viz, multiple DMCL and single DMCL. mDMCL allows
machine rearrangement whereas the latter does not.

Once the demand is being forecasted or being obtained from SMCL covers the data
collection and matrix-based data manipulations along with the application of the formulated
objective functions. Then, pair-wise machines are inserted into the 3x3 space grids in
accordance with the ranking orders. Possible layouts are evaluated and the total part
travelling scores calculated to represent the preferences of the machine layouts. SMCL can
produce a good layout for a single period while DMCL needs a series of static layouts for
multiple planning periods correspondingly, or a particular layout in case of machine
rearrangement is impractical.

In mDMCL, the part travelling cost and the machine rearrangement cost are also to be
considered. The travelling cost can be calculated by converting the travelling score into the
monetary term obtained in SMCL. In period to period machine rearrangement cost consists
of two parts, the basic cost for machine movement and relocation and the cost per unit
machine movement.

The sDMCL tends to keep all machines stationary thus the layout obtained by this way may
not be optimal. The cell layout cost derived from the sDMCL is the aggregation of the part
travelling cost for all the periods. This is done by evaluating every static layout generated in
SMCL to find out the best one, which is going to be used as the final layout for a cell.
Normally, one of them will have the lowest value and this becomes the selected layout of
sDMCL.

3. Layout design model for delayed products differentiation

Manufacturers are no longer interested in standardized products, since markets are neither
homogeneous nor demands for products are stable; consequently, targeting profit from many
market niches became a need, which means increasing products variety through production.
There are many reasons for that instability; such as the existence of different regional
requirements due to cultural or geographically related factors that require a product to have
many models and variants. Even within the same regional market, several distinct market
segments may exist, requiring different functionalities or capabilities for the same product. A
good example for such variation in a product can be found in the automotive industry.
Managing the augmented product varieties is becoming a major concern in manufacturing.
Offering a wide range of product variants leads to a considerable expansion in the number of
sub-assemblies and amounts of raw materials that must be kept in stock to deal with the full
range of possible variations in product configurations. This leads to increased carrying costs
and the risk of stock obsolescence. In addition, customer services cost increases when many
product variations exist.

The potential negative effects of products variety proliferation motivated the establishing of
mitigation methods to manage the resulting complex situation. Adopting a postponement
strategy was a direct response to such a challenge. In manufacturing postponement, raw
materials are shipped to the warehouses at which the final products would be both
manufactured and assembled to reduce transportation and holding costs.

This paper specifically targets the assembly postponement strategy. It requires that product
differentiation activities, which customize the work-in-process inventory into end products,
be deferred as much as possible along production stages by inserting specialized components
or performing special processes. These specialization and customization production stages
are called points of products differentiation.

DPD is considered at two levels:

1. the strategic level, which optimizes the positioning of products differentiation points, and
2. the physical level, which determines the sequence of assembly processes and designs the
physical assembly line layout.

The proposed logical flow of activities in this proposed design framework can be explained
as below.

Product differentiation point: Based on the strategic objectives of the manufacturing system,
such as; investments, production capacity, WIP level, etc., and strategic level parameters
such as inventory holding cost, the production stage at which each product begins to be
unique is established, as well as the level of undifferentiated unfinished products inventories,
which is considered as WIP.

DPD product design rules: Follow design rules such as standardization, modularity,
avoiding precedence constraints, etc.

Commonality Analysis: Use commonality in the products in the family to draw a preliminary
assembly layout fro the given family.

Precedence constraints: This establishes each products assembly sequence and the
geometric and technological validity of the layout.

Feasible assembly sequence: A feasible process sequence for each product either based on
the duplication of the components or relaxing the precedence constraints.
Assembly line layout: a final assembly line layout is being developed including the material
flow pattern and material handling system, and suggesting the positions of the different
assembly process relative to others.

4. Suggestions

By studying both MAIN algorithm and the delayed products differentiation for the layout design
for the layout design we can come to a conclusion that MAIN algorithm tries to optimize the
layout design by considering quantitative variation in the demand where as the delayed products
differentiation deals with the qualitative variation in the demand. At this point considering the
fact that in a manufacturing industry both the quantitative and qualitative variation of the
products is to included in the system. Thus it becomes a difficult job to trade off between
quantitative flexibility and qualitative flexibility.

Here I would like to suggest a different approach to include both the quantitative and the
qualitative flexibility in the system by trying to merge the above two approaches into design
consideration of the layout. This can be tried to explain with an illustrative example as follows

Consider a company producing three components A,B & C. Let these products though different
in design and function, require the same set of machines for production. Or in other words if the
company has five machine all A, B, C can be produced by different combination of these
machines. Now by applying the principles of MAIN algorithm like the total part traveling cost,
the closeness of the machine, the part handling efficiency, the part traveling distance, the
materials flow frequency, and the machine rearrangement cost in the dynamic environment. The
MAIN algorithm can be applied here since the product mix will change in different planning
horizon.

Now we consider the case where these products have slight variations in their design to give
products like A1,A2,A3 for product A. similarly for product B & C, we have B1,B2 & C1,C2,C3
respectively. These variants are does not have large variation in their designs with the general
composite parts A,B& C. So the layout obtained by the MAIN algorithm which will best suit all
the demand in the planning horizon can be fine tuned by applying the principles of the delayed
product differentiation concept. Thus we can try to design a layout which will have the ability to
take both the quantitative and the qualitative demand variation.

But one possible drawback which can be expected while trying to integrate these two concepts is
that some amount of sacrifice has to be made in the range of the demand that can be full filled
and the range of the design variation in the products. But the most important benefit which can
be seen in this case is that this concept of integrating these two concepts and designing a more
robust layout will be beneficial to those products which does not have a very large variation in
there demands or the demand remains more or less within a specified range and the products
which do not have very large variation in the designs.

The feasibility of this suggestion has to be validated by simulation and conclusions have to be
drawn out the results.
Reference:

1. Chan et.al,2004, Development of the MAIN algorithm for a cellular manufacturing


machine layout, IJPR,42: 1, 51-65.
2. AlGeddawy, Tarek, et.al, Assembly systems layout design model for delayed products
differentiation, IJPR,48:18, 5281-5305.

Vous aimerez peut-être aussi