H Ho ow w t to o 0 0s se e m mB BP P I Im m C Co om mt tI Im mu uo ou us s P Pr ro oB Bu uc ct tI Io om m

I Im mB Bu us st tr rI Ie es s W WB Be em m O Or rB Be er r T Ty yg ge e I Is s L Lo ot t L Lo or r L Lo ot t
Ramin Sadeghian`
Ramin Sadeghian, Assistant Professor in Industrial Engineering, Engineering Department, Bu Ali Sina Universitv, Hamedan, Iran,
K KE EY YW WO OR RD DS S ABSTRACT
The Materials Requirements Planning (MRP) method that is applied in
production planning and management has some weaknesses. One of its
weaknesses is that the time in MRP method is discrete, and is
considered as time period. Hence we are not able to order our
requirements at irregular time moments or periods. In this paper, a
new form of MRP is introduced that is named Continuous Materials
Requirements Planning (CMRP) approach. We discuss the
disadvantages of Discrete MRP (DMRP) approach and analv:e the
conditions, applications and the manner of applving CMRP approach
in our problems.
2010 IUST Publication, All rights reserved. Vol. 21, No. 1
1 1. . I In nt tr ro od du uc ct ti io on n
MRP was Iirst introduced in 1970's; thereaIter,
many researchers, papers, books, industries, companies
and even diIIerent sciences have applied it. ( Browne J.
et al., 1996; Horvat L. et al., 1999; O`Neill L. et al.,
2001; Rom W. O. et al., 2002; Dolgui A. et al., 2007).
This system was Iormed as an approach oI production
planning and management. Orlicky (1975) introduced
this system in a help book in 1975. The introduced
MRP was the Iirst version oI MRP system, named as
MRP I. Later, several MRP systems were extended into
other versions including MRP II and ERP. (Browne J.
et al., 1996)
In some researches, the advantages and disadvantages
oI MRPs have been reIerred to. (see Wallace T., 1980;
Latham D., 1981; Hinds S., 1982; Burbidge J., 1985;
SaIizadeh M. et al., 1986; Harhen J., 1988; Browne J.
et al., 1996). In this paper, some other advantages and
disadvantages oI this system will be pointed out.
Almost in all oI related researches, MRP system has
been considered as a DMRP; since orders, demands,
Corresponding author. Ramin Sadeghian
Email: rsadeghianbasu.ac.ir
Paper first received March. 10. 2010 ,and in revised form 1une.
30, 2010.
scheduled receipts and etc, are deIined in discrete time
or distinguished time periods. (Enns S. T., 1999;
Johnny C. Ho. et al., 2001; Gutierrez Y. B. et al., 2008;
Pillai V. M. et al., 2008)
In this paper, we explain the weaknesses oI a DMRP
system and the manner oI applying CMRP method in
some problems. In the Iollowing sections, some needed
deIinitions and the manner oI modeling CMRP are
pointed out and some problems related to DMRP are
solved.
Note that all over this paper, continuous production
means a production with continuous time and not
continuous supply or demand, or DMRP can also be
applied when supply and demand are continuous.
CMRP is preIerred over DMRP, because DMRP can
not be applied in the industries oI Petrochemical, Gas,
Oil, Water and Waste and other continuous production
industries, while CMRP can be. Also in CMRP
approach, instead oI Gross Requirements (GR),
Scheduled Receipts (SR), On Hand (OH) inventory and
other parameters applied in discrete Iorm, some
continuous Iunctions such as regression Iunctions,
interpolations or extrapolations and even multi rules
Iunctions can be deIined, which enables us to perIorm
sensitivity analysis and Iorecast the required
parameters oI the model.
Discrete Materials Requirements
Planning (DMRP),
Continuous Materials
Requirements Planning (CMRP),
Order Type (OT),
Lot Ior Lot (L4L)
1 1u un ne e 2 20 01 10 0, , V Vo ol lu um me e 2 21 1, , N Nu um mb be er r 1 1
I II I I II In nn n n nn ni ii i i ii ie ee e e ee er rr r r rr rn nn n n nn na aa a a aa ai ii i i ii ii ii i i ii io oo o o oo on nn n n nn na aa a a aa aI II I I II I ] ]] ] ] ]] ]o oo o o oo ou uu u u uu ur rr r r rr rn nn n n nn na aa a a aa aI II I I II I o oo o o oo of ff f f ff f I II I I II In nn n n nn nd dd d d dd du uu u u uu us ss s s ss si ii i i ii ir rr r r rr ri ii i i ii ia aa a a aa aI II I I II I I II I I II In nn n n nn ng gg g g gg gi ii i i ii in nn n n nn ne ee e e ee ee ee e e ee er rr r r rr ri ii i i ii in nn n n nn ng gg g g gg g & && & & && & I II I I II Ir rr r r rr ro oo o o oo od dd d d dd du uu u u uu uc cc c c cc ci ii i i ii ii ii i i ii io oo o o oo on nn n n nn n I II I I II Ie ee e e ee es ss s s ss se ee e e ee ea aa a a aa ar rr r r rr rc cc c c cc cn nn n n nn n
1 1o ou ur rn na al l W We eb bs si it te e: : h ht tt tp p: :/ // /I I1 1I IE EP PR R. .i iu us st t. .a ac c. .i ir r/ /
I In nt te er rn na at ti io on na al l 1 1o ou ur rn na al l o of f I In nd du us st tr ri ia al l E En ng gi in ne ee er ri in ng g & & P Pr ro od du uc ct ti io on n R Re es se ea ar rc ch h ( (2 20 01 10 0) ) p pp p. . 1 17 7- -2 22 2
ISSNt 2008-4889
1 18 8 Ramin Sadeghian H Ho ow w t to o U Us se e M MR RP P i in n C Co on nt ti in nu uo ou us s P Pr ro od du uc ct ti io on n I In nd du us st tr ri ie es s W Wh he en n O Or rd de er r 1 1y yp pe e i is s L Lo ot t F Fo or r L Lo ot t
2. Definitions
In DMRP approach, generally, the Iollowing
notations are used, which may vary in some reIerences
(Browne J. et al., 1996; Gutierrez Y. B. et al., 2007):
i
GR : Gross Requirements during time period i
i
SR : Scheduled Receipts during time period i
i
OH
: On Hand inventory during time period i
i
NR
: Net Requirements during time period i
i
PO
: Planned Orders during time period i
where GR and SR, during all time periods, and OH, in
the beginning oI planning horizon, are known, the
unknown parameters can be obtained by the known
ones.
For obtaining the parameters in DMRP approach the
value oI
T
SR SR OH , , ,
1 1
TT T GR GR
T
, , , ,
1

should be speciIied, where the planning horizon is

|1,T|, TT is operations' total time Ior each piece oI
product when is deIined as sum oI Iour types oI time
consisting oI queue time (QT), process time (PT),
carrying time (CT) and setup time (ST) as Iollows:
ST CT PT QT TT + + + =
II OT is Lot Ior Lot so:
i i i i
OH SR GR NR = (1)
i i i i i
GR SR OH NR OH + + =
+1
(2)
TT i i
NR PO
+
=
All oI the above deIinitions and notations are only used
in DMRP approach. In CMRP approach they are
deIined as Iollows:
) (t GR : Gross Requirements in time moment t
) (t SR : Scheduled Receipts in time moment t
) (t OH
: On Hand inventory in time moment t
) (t NR : Net Requirements in time moment t
) (t PO : Planned Orders in time moment t
3. Modeling
Assume that
T
GR GR GR , , ,
2 1
are gross
requirements in time periods 1, 2, ..., T, so ) (t GR
Iunction can be obtained in the Iorm oI a regression
Iunction by
T
GR GR GR , , ,
2 1
. ) (t GR can be a
polynomial, exponential or any other type oI Iunctions.
Some assumptions are considered in this paper as
Iollows:
- TT is constant and known
- OT L4L
Similarly, ) (t SR can be deIined by the data oI past
periods in the Iormat oI a continuous Iunction.
Lemma 1: In CMRP approach, iI OT and SS are equal
to L4L and zero respectively, so ) (t OH can be
determined by the Iollowing equation:

+
=


Otherwise
p i t t t dt t GR dt t SR
t t t dt t GR dt t SR t OH
t OH
i i
t
t
t
t
t
t
t
t
i i
0
} , , 2 , 1 ; ) ( ) (
) ( ) ( ) (
) (
*
*
0 0 0
0 0

(3)
where
0
t is the start time oI planning horizon and
*
0
t is
a point that
0 ) ( ) ( ) (
*
0
0
*
0
0
0
= +

t
t
t
t
dt t GR dt t SR t OH
, and
| , | , ,
0 1
T t t t
p
are the points in which
) ( ) (
i i
t GR t SR =
and 0 ) ( ) ( ; 0 > + + >
i i
t GR t SR ,
also | , | , ,
0
* *
1
T t t t
p
are the points that
} , , 1 ; 0 ) ( ) (
* *
p i dt t GR dt t SR
i
i
i
i
t
t
t
t
=

.
Also
* *
1
*
0
, , ,
p
t t t should be minimum solution obtained
Irom their corresponding equations.
Proof:
According to Eq. 2, OH in DMRP approach can be
written as Iollowing:
T i GR SR OH NR OH
i i i i i
, , 1 ;
1
= + + =
+
It is clear that, while
i i i
GR SR OH OH + > , 0
1
,
0 =
i
NR . In such a case the above equation can be
rewritten as Iollows:
T i GR SR OH OH
i i i i
, , 1 ;
1
= + =
+
Now, use returnable equations Ior the above equations
as Iollows:
Ramin Sadeghian H Ho ow w t to o U Us se e M MR RP P i in n C Co on nt ti in nu uo ou us s P Pr ro od du uc ct ti io on n I In nd du us st tr ri ie es s W Wh he en n O Or rd de er r 1 1y yp pe e i is s L Lo ot t F Fo or r L Lo ot t 19

= =
+
+ =
= + + + + + =
= + =
+ + + =
= + + + =
= + =
+ + =
= + + =
= + =
+ =
i
f
f
i
f
f
i i
i i i i
GR SR OH
GR GR GR GR SR SR SR SR OH
GR SR OH OH
GR GR GR SR SR SR OH
GR SR GR GR SR SR OH
GR SR OH OH
GR GR SR SR OH
GR SR GR SR OH
GR SR OH OH
GR SR OH OH
1 1
1
3 2 1 3 2 1 1
1
3 2 1 3 2 1 1
3 3 2 1 2 1 1
3 3 3 4
2 1 2 1 1
2 2 1 1 1
2 2 2 3
1 1 1 2

ThereIore,

= =
+
+ =
i
f
f
i
f
f i
GR SR OH OH
1 1
1 1
(4)
All relations oI DMRP approach, with some subtle
changes, can be used Ior CMRP approach. For
example, summation notations are transIormed into
integral notations.
With respect to the above comments, the Eq. 4 Ior
CMRP approach can be rewritten as Iollows:

+ =
t t
dt t GR dt t SR OH t OH
0 0
) ( ) ( ) 1 ( ) (
II the beginning point oI planning horizon is
0
t , then:

+ =
t
t
t
t
dt t GR dt t SR t OH t OH
0 0
) ( ) ( ) ( ) (
0
(5)
The above equation is correct iI
) (
0
t OH
and
cumulative ) (t SR are able to satisIy cumulative
) (t GR , so iI
*
0
t is the point that
0 ) ( ) ( ) (
*
0
0
*
0
0
0
= +

t
t
t
t
dt t GR dt t SR t OH
, and
*
0
t is
minimum solution obtained by the corresponding
equation. It means that Ior the points aIter
*
0
t ;
) (
0
t OH and cumulative ) (t SR are not able to satisIy
cumulative ) (t GR , hence; Eq. 7 is correct only Ior
interval
*
0 0
t t t .
Similarly, assume | , | , ,
0 1
T t t t
p
are the points in
which ) (
i
t SR is equal to
) (
i
t GR
, so the previous
paragraphs' comments are correct when
0 ) ( ) ( ; 0 > + + >
i i
t GR t SR . In this case
) (
i
t NR
will also be equal to zero, the related equation
in discrete case is

= =
+
=
i
f
f
i
f
f i
GR SR OH
1 1
1
,
because
1
OH has already been consumed completely,
and in continuous case it can be written as Iollows:

=
t
t
t
t
i i
dt t GR dt t SR t OH ) ( ) ( ) ( (6)
For determining the mentioned intervals, it is suIIicient
to Iind the points
* *
1
, ,
p
t t , subject to
} , , 1 ; 0 ) ( ) (
* *
p i dt t GR dt t SR
i
i
i
i
t
t
t
t
=

, and
* *
1
, ,
p
t t
are minimum solution obtained Irom their
corresponding equations. Thus, Eq. 6 is correct during
the intervals p i t t t
i i
, , 1 ;
*
= .
On the other hand, since OT is L4L, in the points that
) (t NR is not zero, the quantity oI ) (t NR is exactly
equal to the requirements oI system, hence; there is not
any extra inventory Ior stocking. Consequently in these
points ) (t OH is equal to zero.
ThereIore; the prooI is completed.
) ( ), ( t PO t NR can also be determined by ) ( ), ( t SR t GR
and ) (t OH in any moment oI time as Iollows:
) ( ) ( ) ( ) ( t SR t OH t GR t NR = (7)
) ( ) ( TT t NR t PO + =
In DMRP approach, TT should be integer, while in
CMRP approach TT can also be considered as a real
number. This Iact is the other advantage oI CMRP
approach over the DMRP one.
2 20 0 Ramin Sadeghian H Ho ow w t to o U Us se e M MR RP P i in n C Co on nt ti in nu uo ou us s P Pr ro od du uc ct ti io on n I In nd du us st tr ri ie es s W Wh he en n O Or rd de er r 1 1y yp pe e i is s L Lo ot t F Fo or r L Lo ot t
Lemma 2: In a CMRP system, whose OT is equal to
L4L and its SS is zero, ) (t NR can be determined by
the Iollowing equation:

=
Otherwise t SR t GR
p i t t t
t NR
i i
) ( ) (
} , , 2 , 1 , 0 ; 0
) (
*

(8)
where
0
t is the start time oI planning horizon and
*
0
t is
a point that
0 ) ( ) ( ) (
*
0
0
*
0
0
0
= +

t
t
t
t
dt t GR dt t SR t OH
, and
| , | , ,
0 1
T t t t
p
are the points in which
) ( ) (
i i
t GR t SR =
and 0 ) ( ) ( ; 0 > + + >
i i
t GR t SR ,
also | , | , ,
0
* *
1
T t t t
p
are the points that
} , , 1 ; 0 ) ( ) (
* *
p i dt t GR dt t SR
i
i
i
i
t
t
t
t
=

. Also
* *
1
*
0
, , ,
p
t t t
should be minimum solution obtained Irom their
corresponding equations.
Proof:
It is similar to the prooI oI lemma 1.
4. Numerical Example
Example: Suppose that Iunctions
20 4 2 . 0 ) (
2
+ + = t t t GR
and 40 5 2 . ) (
2
+ = t t t SR ;
| 20 , 1 | t
represent gross requirements and scheduled
receipts in time moment t respectively; moreover,
10 ) 1 ( = OH
and TT1.5 time units. Then, by algorithms I
and II the Iunctions oI ) ( ), ( t NR t OH and
) (t PO
can
be determined as Iollows:
0 634 . 5 20 5 . 4 134 . 0
0 20 4 2 . 40 5 2 . 10
0 ) ( ) ( ) 1 (
1
2 3
1
2
1
2
1 1
0
= +
= + + + +
= +
=


t t t
t t t t
t GR t SR OH
t
t t
t t
906 . 4
| 20 , 1 | 374 . 28
906 . 4
| 20 , 1 | 302 . 0
*
0
*
*
*
=

=
=
=

t
Refected t
Accepted solution minimum t
Refected t

=
=

= +
+ = + +
=
20
5 . 2
0 20 9 4 . 0
40 5 2 . 0 20 4 2 . 0
) ( ) (
2
1
2
2 2
t
t
t t
t t t t
t SR t GR
) ( ) ( ; 0
1 1
+ >/ + > / t GR t SR and 20
2
= t is the
end point oI planning horizon, thus calculating
*
1
t and
*
2
t is not necessary. Hence; the Iunctions oI
) ( ), ( t NR t OH and
) (t PO
are obtained as Iollows:

+
=
Otherwise
t t t t
t OH
0
906 . 4 1 634 . 5 20 5 . 4 134 . 0
) (
2 3

+

=
Otherwise t t
t
t NR
20 9 4 . 0
906 . 4 1 0
) (
2

+

=

+ + +
+
= + =
Otherwise t t
t
t PO
Otherwise t t
t
t NR t PO
6 . 15 6 . 8 4 . 0
406 . 4 1 0
) (
20 ) 5 . 0 ( 9 ) 5 . 0 ( 4 . 0
906 . 4 5 . 0 1 0
) 5 . 0 ( ) (
2
2
In the Iirst rule oI PO(t) Iunction, t can be placed in the
interval |0.5,4.406|. Since the planning horizon is
|1,20|,the mentioned interval is written as |1,4.406|.
For comparing the obtained Iunctions and their curves,
see Iigures 1 to 4: (Note that in all charts oI this paper,
the horizontal axis shows time unit and the vertical axis
shows the speciIied parameters in each chart.)
Fig. 1. The curves of GR and SR
Fig. 2. The curves of GR and NR
Ramin Sadeghian H Ho ow w t to o U Us se e M MR RP P i in n C Co on nt ti in nu uo ou us s P Pr ro od du uc ct ti io on n I In nd du us st tr ri ie es s W Wh he en n O Or rd de er r 1 1y yp pe e i is s L Lo ot t F Fo or r L Lo ot t 21
Fig. 3. The curves of NR and OH
Fig. 4. The curves of NR and PO
A signiIicant advantage oI CMRP approach is that it
enables us to determine the required amount oI product
or order in each time interval or moment. For example,
iI the present time is
1
t , the required quantity oI
product during interval | , |
2 1
t t is determined by the
Iollowing equation:
Required quantity oI product during interval | , |
2 1
t t

2
1
) (
t
t
dt t NR
Also the required quantity oI order during interval
| , |
2 1
t t can be determined by the Iollowing equation:
Required quantity oI order during interval | , |
2 1
t t

+ =
2
1
2
1
) ( ) (
t
t
t
t
dt TT t NR dt t PO
While, the required quantity oI product or order in any
arbitrary time interval can not be ever determined by
DMRP approach.
5. Conclusion
In this paper, CMRP approach was introduced and
some advantages oI this approach over DMRP one
were explained. Three lemmas with their prooIs and
three algorithms were also described. Later, two
numerical examples were solved completely. Some
priorities oI CMRP approach over DMRP one are as
Iollows:
- DMRP can not be usually applied in some
industries such as Petrochemical, Gas, Oil, Water
and Waste and other continuous production
industries, while CMRP can be.
- In CMRP approach, Ior GR, SR, OH and other
parameters, some continuous Iunctions can be
deIined, whose applications enables us to perIorm
sensitivity analysis and Iorecast the required
parameters oI the intended model.
- In DMRP approach, TT have to be integer,
while in CMRP, TT can also be a non-integer
number.
References
|1| Browne, J., Harben, J., Shivnan, J., Production
Management Svstems. an Integrated Perspective,
Second Edition, Addison-Wesley Press, 1996.
|2| Burbidge, J., Automated Production Control, Modelling
Production Management Svstems, Amsterdam, North-
Holand. 1985.
|3| Dolgui, A., Prodhon, C., Supplv Planning Under
Uncertainties in MRP Environments. A State of the Art,
Annual Reviews in Control 31, 2007, pp. 269279.
|4| Enns, S.T., The Effect of Batch Si:e Selection on MRP
Performance, Computers & Industrial Engineering 37,
1999, pp. 15 -19.
|5| Gutierrez, Y.B., Diaz, B.A., Gupta, S.M., Lot Si:ing in
Reverse MRP for Scheduling Disassemblv, International
Journal oI Production Economics 111, 2008, pp.741
751.
|6| Harhen, J., The state of the art of MRP/MRP II, Computer
Aided Production Management. The State of the Art,
Springer-verlag, Germany, 1988.
|7| Hinds, S., The Spirit of Materials Requirements Planning,
Production and Inventory Management, 23(4), 1982, pp.
35-50.
|8| Horvat, L., Bogataj, L., A Market Game with the
Characteristic Function According to the MRP and
Input-Output Analvsis Model, International Journal oI
Production Economics 59, 1999, pp. 281-288
|9| Johnny C. Ho., Yih-Long Ch., An Integrated MRP and
JIT Framework, Computer & Industrial Engineering 41
2001, pp.173-185.
|10| Latham, D., Are You Among MRPs Walking Wounded?,
Production and Inventory Management, 22 (3), 1981,
pp. 33-41.
|11| O`Neill, L., Murphy, M., Gray, D., Stoner, T., An MRP
Svstem for Surgical Linen Management at a Large
Hospital, Journal oI Medical Systems, Vol. 25, No. 1,
2001.
2 22 2 Ramin Sadeghian H Ho ow w t to o U Us se e M MR RP P i in n C Co on nt ti in nu uo ou us s P Pr ro od du uc ct ti io on n I In nd du us st tr ri ie es s W Wh he en n O Or rd de er r 1 1y yp pe e i is s L Lo ot t F Fo or r L Lo ot t
|12| Orlicky, J., Materials Requirements Planning. The New
Wav of Life in Production and Inventorv Management,
New York, McGraw-Hill, 1975.
|13| Pillai V. M. and Chandrasekharan M. P., An Absorbing
Markov Chain Model for Production Svstems with
Rework and Scrapping, Computers & Industrial
Engineering 55, 2008, pp. 695706.
|14| Rom, W.O., Tukel, O.I., Muscatello, J.R., MRP in a Job
Shop Environment Using a Resource Constrained
profect Scheduling Model, Omega 30, 2002, pp. 275
286.
|15| SaIizadeh, M., RaaIat, F., Formal/Informal Svstems and
MRP Implementation, Production and Inventory
Management, 27 (1), 1986.
|16| Wallace, T., APICS Dictionarv, Fourth Edition,
Washington, DC, American Production and Inventory
Control Society, 1980.