284
mw mK #
a ikl | =
an a
wan A Ae. (ait. Hed. HH) RHRT
Application of System Dynanies in R&D Praject Plaon
ches Aisha HALA HA fA if
WARE AR. APM, RAtK, BEE, He
BRC19 81316 SEXAKT “HERREEAH” 8
polar maeye 227927 Rankes, Morweig ian Shoe! of Managent
Hea “Baraise Ver307 138 Bakkes tia, Norway
Sh gat hi apt Jude fia » WEGT.
& &
J
*
wi
HR HAR.
APPLICATIONS OF SYSTEM DYNAMICS IN R&D
PROJECT PLANNING AND POLICY ANALYSIS
Ching-rui Xu
Dept. of Industrial Engineering
Zhejiang University
Hangzhou, China
ABSTRACTS
Emphasis on economic effectiveness and benefit in our
country results in the emphasis of economic analysis in pro-
ject planning and evaluation during recent several years,
The weakness of existing approaches in project planning
and analysis to certain extent is lack of dynamic in nature,
The effectiveness of system dynamic approach in project
planning and analysis is not only due to its systematic and
dynemic analysis, but also due to its value in quantitative
analysis and policy analysis. The idea and model of R&D pro-
ject planning is useful in solving above-mentioned problems,
The learning curve nature in development activities,
Adoption of task-performance coefficient as a factor in
R&D system dynamics modeling.
The labor psychological factor in our country and its
characteristics in formulation of system dynamics simulation
model,
R&D cost as a major element is involved in the model.
Policy analysis through simulation running is an impor-
tant basis for decision-making in R&D project planning.
In 1950s, our country began to emphasize the economic
analysis of research and development projects in the large
scope of economic construction, Significant economic benefits
were achieved, Therefore, it played a large part in achiev-
ing better benefits of investment and better economic effects
both on the society as a whole and on enterprises and in spee-
ding up the development of our socialist construction.
285
THE ECOMONIC BFNEFIT OF SOCLALISN FIRST OF ALL CONLS FRON
PLANNING
The experience of economic construction in the First
Five-Year Plan and the early three years of the Second Five-
Year Plan period in our country proved that stressing the
economic benefits was e crucial principle, especially to a
developing country, Capital for investment is the scarcest
resource in a developing country. This is also true to our
nation, There are many ways to solve this problem, As to our
nation, we must rely heavily on self-reliance. Of course, it
does not mean that we have to shut the door against every
other nation, We can communicate with developed countries in
science, technology and economy, including the use of techno-
logy and capital from foreign country, However, the main sour-
ce of capital for construction is domestic ones. Since there
is a vi
t domestic market in our nation, it becomes an im-
portant factor in ensuring for our country steady increase
of economy without being interrupted by the economic crises
throughout the world.
In order to base ourselves on home when building up the
capital for the socialist economic construction in our nation
we should strees the economic benefit, bring economic acti-
vities on the course of achieving better economic results and
make economic evaluation of such projects as research, deve-
lopment and planning ones. Only when focusing on it, can we
have correct and effective decision-making,
3.
Since our country is a socialist one, the primary and
fundawental characteristic of a socialist economy should be a
planned system, There should be a overall national economic
plan, in the light of which science, technology, production,
transportation and so on are smoothly managed and integrated.
The economic results of. work performance depends, to a
great extent, on the quality of planning and wether the plan-
ning itself implements the guideline of achieving better econo-
mic results, Moreover, China being a socialist nation, all
plans, ranging from enormous macro-plans such as of national
economy to small micro-plans such as of enterprises as well
as of project prograns, should carry out the guideline of achie-
ving better economic effects. In the process of planning, the
economic evaluation should be made, Similarly, in making plans
of science and technology, including R&D plans and the plans
of technology progress, we should undoubtly strees technolo
gical and economic evaluation, striving for better economic
results and make every effort to achieve the best economic re-
sults with the least manpower and the fewest material resources,
ECONOMIC EVALUATION OF A R&D PROJECT AND ITS PLANNING
In a quite long period we followed the theory and method
of technological and economic analysis introduced from the
Soviet Union. There were some serious shortcomings in this
approach, the main one of which was the lack of overall dynamic
analysis, It was related to the limitation of the theories and
methods of management and economic analysis in that period
ode
(1950s). It seemed to have seriously damaged the planning,
especially to R&D planning. Because the crux of planning lay
in overall and comprehensive balance and in long and middie
range prediction for the future, failing to analyse the pro-
blems comprehensively and lacking the prediction can hardly
assure the correctness anu effectiveness of planning and de-
cision-making. R&D planning was characterized as long-range
one. Any approach without involving dynamic analysis for long-
range appeared unable to survive. Consequently, its effecti-
veness would be drastically reduced.
Another shortcoming of the existing approach was its
inability of utilizing modern approaches and means effective-
ly, especially its inability of applying computer-aided deci-
sion-making techniques effectively. As a result, it cannot com-
pare a large number of schemes and employ quantitative tech-
niques to analyse all alternative policies in order to seek
out the optimum decision as a basis for policy making.
System dynamics displayed significant superiority in sol-
ving above-mentioned problems. It can analyse R&D projects
systematically, dynamically and quantitatively, In addition,
it can analyse R&D projects and select the best ones, when
multi-indices exist and a lot of projects are compared simul-
tanously,.
SYSTEM DYNAMIC NODELS OF R&D PROJECT PLANNING
It is necessary to distinguish the research activities
286
Se
from development ones and make plans and establish system dy-
namic models according to their own distinctions. The practice
of our nation has indicated that the development activities
are more determinate and are able to formulate task norm. Thus,
quantitative techniques and system dynamic models can be effe-
ctively employed when particular planning of development pro-
jects is made.
This article employed the R&D model set up by Edward
Roberts as a basis to discuss the problems and relate the expe-
rience of our nation to probe into the problem of R&D modeling
as well as decision-making in this area.
Following problems are. worth clerifing:
* The learning curve effect in R&D project.
* Motivation and its effectiveness in R&D project.
* The attitude and psychological factor in R&D project com=
pletion.
* Use of comprehensive index-system for R&D project evalua-
tion and choice.
The model to be presented consists of wore than fourty
equations (see appendix). It is composed of four parts: (1) Pro-
gress; (2) Productivity; (3) Manpower; (4) Change in Cost. The
first part is involved in the real progress rate, level of cum-
mulative real progress etc. This part is similar to the work
done by Professor Ed. Roberts, The 2nd part is concerned with
productivity, two characteristics are appearently appeared in
our country. One is the learning curve effect in development
287
6,
work, Development work, including machine design, is different
from those of research work, it is an engineering work, a com-
bination of uncertainty ond routinized work with much more cer~
tainty than research work, According to the practice of some
machine-tool manufacturing firms in our country, a learning
curve effect could be perceived, which is to some extent simi-
lar to the producion one, (See Fig.1)
Pig.1 is 9 curve
concerning design time
of machine tool compon-
ents (parts) over time.
For instance, in a ma-
chine-tool company,
which initiated to de-
sign & produce grinding
machine in 1953, the
average design time of
one equavalent compon-
ent of the machine tool
was changing over years.
ear
1953
1957
‘.
!
i)
i H
q 1
Py eC
Fig. i
design time required
14 hours or more
hours
This factory develops more than ten kinds of new grinding
machine e
ch year, which offers the designers an opportunity to
raise their skill and productivity, By the time this machine
+7
tool factory also gives out a reward system for engineer and
designers.
According to this learning curve effect in development,
it is suitable to add "Norm Performance Factor" (NPF), by use of
CLIP Function in the "Productivity" part of our model as
follows,
PROD. K=(PM.K) (PRODN)(NPF.K)
PRODN=2
PM.K=TABLE(TPR, RFSC.K, 0, 2, .25)
TPM= .6/.65/.75/.85/1/1.3/1.5/1.6/1.45/1/.8
NPF.K=TABLE(TNPF, TINE.K, 0, 72, 6)
TNPF= .6/.7/.8/.85/.9/.95/1/1.02/1.04/1.055/1.065/1.075/1.082
here, NPF---Norm Performance Factor
(skill from Practice over time)
CLIP---Clip Function
PN---Productivity Multiplier
PRODN---Productivity, Normal(Job unit/man-month)
TPN--~Teble Productivity Multiplier
RFSC---Ratio of Forecast to Scheduled Completion Dates
In 1950s, many of our mathine-building manufactures initiate
a@ reward system in R&D Dept by use of norms (or work standard)
for developing work and design work. A variety of norms is iden-
tified to different kinds of jobs, euch as design of machine com-
ponents, design of process, design of tool used in plant for pro-
ducing new components and so on, The norm is also classified
according to the difficulties and novelty of the machine. These
288
Be
norms are periodically reviewed and modified by use of statisti-
cal and empirical data. A reward system wes well designed to fit
the nornm-performance system of designing and developing work.
In these several years, Economical Responsibility System has
been erected and widely adopted in our industry. And, the above-
mentioned norm-performance-reward system is integrated with the
economical responsibility system, and plays a great role in rais-
ing the productivity of R&D activities in our factory.
For instance, according to the practice of Shanghai Ma-
chine-Tool Company, since the norm-performance-reward system
was. integrated with the economical responsibility system, the
performance of norm(work standard) for design work is 15-20%
higher than before, in other words, the design time of a com-
ponent. is reduced by 15-20; than the original.
NPF
. Yr.
Fig. 2. Norm-Rerformance Factor
Curve
oe
Based on the learning curve effect, we can draw the curve
of Norm-Performance Factor as shown in Fig.2
Corresponding to the Norm-Performance Factor curve, we
may draw out a series of norm-performance coefficients as
below: .
TNPF=.6/.7/.8/.85/.9/1/1.02/1,04/1.055/1.065/1.075/1,082
Certainly, these series of coefficient would influence
to behavior of productivity as well as that of real progress
of the project.
Another important character having to be discussed here is
about the productivity multiplier (PM). As known, the actual
productivity of the average engineer/scientists working on the
R&D job is not merely influenced by the increasing skillfulness
and sophistication in the appearance of the learning curve, but
also influenced by the attitude of the engineer/scientists and
the motivation, as well as the pressure on the schedule. All
these factors impact the productivity through the productivity
multiplier, In this side, the social and psychological aspects
as well as the culture play a great role, The people of our na-
tion is diligent and will not be frightened by any difficulty.
Our motto is "The more difficulty there is, the more action we
will take" "Difficulties can never scare us". Especially, under
the pressure of schedule and urgent mission, usually the engi-
neer and scientist can perform the work standard over 50-80%,
sometimes 100% or more, Due to all of these, the productivity
multiplier curve would be in different manner from the western
289
210,
Fig. 3), Me
style in average (see Fig. 3) 15 WEPMM---Wage Plus Expenses per Man-Nonth (Yuans) y
Here, only gives the idea, a more complete part of cost is needed
to be developed further.
POLICY ANALYSIS---SYSTEM BEHAVIOR
The model presented. here is only an approximation of the
complex system of research and development projects. However, the
Rodueburty Multipher
eres sRrre
‘system characteristics involved are sufficiently broad that they
may be used for policy analysis based on DYNAMO simulation results.
. LL L (1) BASIC NODEL BEHAVIOR
.
Py ines d o ched There are several key dynamic variables during the project
Pa fio. OF st Yo
Completion Date life as shown on Fig.4, The project initiated at zero time with
Fry. 2
planned completion time of 30 months. Assuming a normal producti-
Corresponding to the PM Curve in Fig.3, we can obtain a se- vity of 2 units per man-month, the project effort is 600 man-month
ries of PM data as follows:
of R&D work. If we spread this effort evenly over 30 months, it
TPMa .6/.65/.75/085/1/163/165/1.6/1.45/1/.8
will require 20 engineers/scientists on the project.
For the sake of taking a whole view in reviewing and asses- dindoiiGa SWALRL. COALEIOA, Wen eeags GERAUTLINTEy oF
sing a R&D project planning decision, it is necessary to include person is less than 2 units, A basic problem in R&D is the rela-
some indices in R&D expense dimension, Accordingly, a fourth part
tive intangibility for most of the work, particularly during the
of cost is arranged in the model as below: early phases of a research project. Because of the intangibility,
CMEN .K=CNEN . J+ (DT) (MEN.J) in general the perceived (and scheduled) progress, based on 2 units
CMEN=0 per man-month, cumulates at a fa
er rate than the actual progress
CCOST, K=WEPMM*CNEN .K (shown as the "A" curve in Fig.4). and, this formed gap is not de-
WEPMM=1980 YUANS tected until month 18 in the simulation, from that date changes
CMEN--~Cumulative Men on Project (Man-Nonth) begin in project behavior. Two observable changes would occur.
CCOST---Cumulative Cost (Yuans)
First, under schedule pressure due to the deviation of fo-
40.000
0.7500
20,000 30.000
0.5000
0.2500.
10.000
0.0000
290
ete +13.
gse
ane WTP US SETS BT eee: Sco ay BE JS IS 8 S&B) acd] ceues e recast completion from schedule, productivity of R&D persons
“34 1
Fei !
sae 1 ‘ ' begin to rise, peaking at nearly 2(more than 1.9) job units
ehalalaieheh state tetas i ¥
Fi he 7 > poe few per man-month,i.e, about an increase of 20.) than the produc-
t | Pereeived 1 fr i tivity of early period. This rising productivity gradually
' 1 : 1 >
ges | Gimalative » : drops down as the gap between forecast and schedule getting
BRS... .. 2...) Rragress, fF. !
ols, ‘ a gl a smaller and smaller,
Tam ' >
' ' : > >>>
woos 1 a i we 1 The second change is that the firm assigns wore R&D work
' t a cS
' 1 of 20 men up to
1 vd force to the project, going from initial leve
\ Productivity $ fe
eos 25 persons at nearly the end of the project completion date,
S88 |
1 :
Bis These changes result in project completion during month
Hak eB
1
i 33, 10 percent slippage of the original schedule. The total
! effort required is 709 man-months, in contrast to the ideal
iz case of 600 man-months. Be sure that the increasing producti-
208 ‘
sks vity starting about month 19 does give benefit to the comple-
cated
Ba8! ae t 1 tion of the project. This productivity change thus produces
’ 1 i
' j
' L« 1 1 fe approximately 5 percent saving of total effort, hence, total
1 Pe * Cos
i pis ' ' cost in. the, project.
.' £e i ' i
gees ' 1 (2) PLANNING INPROVENENT---ACCURATE PROGRESS PERCEPTION
SOO a ee 6 oe RIE, ik
BS 6 = shen oma eas & The serious problem discussed in the preceding section
Nos 3 g 2
g ° § 3 is that lack of tangibility results in delay until the 18th
a é q E
o “a S o
o 5 month in the recognition of project problems, If we take the
MONTHS _—_—_——
y . policy of improving planning and the measure of accurate pro-
MENSM = AMEN=R = PUCSA PP UC=F scone:
PPROD=Y CCOST=* SCOM=$ | FCD=F so pPROD=x gress perception,i.e, any error in perceived progress is imme
Fig.4 Basic Model Simulation ae diately detected and corrected. Under this policy, an expe-
riment of project simulation is wade and illustrated in Fig.5,
40.000
0.7500
40.000
1.7750
30.000T
Productivity
x
30.000
0.5000
35.000
1.4500
20.000T
20.000
0.2500
30.000
1.1250
10,000T
20,000 - ---- -f- 2 -- -
al
Al
yo"
~
ap
10.000
0.0000
25,000
0.8000
0.000T
0.0000 A
10.000 - - -
30.000 - - - - - -
MONTHS —__ |
PUc=A
MENEM AMEN=R
PPROD=Y CCOST=*
PPJC=P SCOM=S
FensF pron=xi
Fig.5 Plan Improvement: Accurate Progress Perception
291
215.
here perceived actual progress remain together throughout the
project, Cumulative required effort is 722 man-months, slightly
more than those in the basic model simulation, but other bene-
fits results from the accurate progress perception. The project
is completed during month 32, 2 months behind initial schedule,
but 1 month ahead of fore-mentioned basic case, Furthermore,
peak manpower is 24 R&D persons instead of 26 of the basic case,
which creates a certain extensive improvement in stability of
the organization. These results, however, are not significantly
different from the earlier basic case.
Policies for managing R&D projects give significant in-
fluence on the results, The next three policies are those rela-
ted to schedule and R&D work-force changes,
(3) SCHEDULE-FIRST POLICY
According to this policy, the completion following to
schedule is the first-of-all policy, in which the initially re-
gulated schedule is treated as fixed rather than flexible. This
ion
policy is employed under the circumstances of urgent mi
subsyst.
to be accomplished, of short-line or “bottle neck"
within a complex project of large-scale system, as well as in
the situation of "crash" project and many other R&D project in
which the time of completion is given the highest priority. The
simulation results under this policy shown in Fig.6 demonstrate
that as the forecast completion date rises in response to recog-
nition of errors in progress perception, the scheduled comple-
tion date is held nearly fixed at its initial regulated period.
292
essce Several principal change results: +17,
$88ssi
3i8 us! * The productivity rises siguificently, peaking at more than
'
oe 2 units per man-nonth during 27th wonth, an increase of
a.
; i
f 1 Rerceived ' near 40 percent over initial productivity.
' f hi
lative
1 1 Camale ' * The manpower level on the project is greatly increased,
is
ee9og! 1 f °
gssuse Me eae * rising up to more than 31 persons, i.e, an increase of
SBSRS1 eee eee ee r
Be he :
SF a8! ' = i more than 50 percent over the original level of 20,
r t
ct
' Productvify of > . ' * The cost level on the project is highly raised, increa-
f i
i : = >» sing up to more than 35 percent over the basic case and
! \ «forecast > he si ai d thi 2
“ ‘ the situation under other policies.
gegesi f Compkton |
S8SBSi eee Pret eee are ‘Date’ gr * The project is completed by the month 31, a slight delay
seats 1 "
RoR AR! LY ' Z _ AF of 1 month behind the original schedule.
7 2
ul t '
it 4 t (4) FIXED-MANPOWER POLICY
5 '
ft i Under this policy no work-force change is made to respond
Ld n i
883 83 ‘ Oo to small change in schedule. This policy is usually used for
Ce we ; ;
eos a8! : ' basic research work at early period, or for no time limitation
1 ' ss ! project, The result of simulation shown in Fig.7, indicate:
1 '
i vat i 1 Gst ' * Give no changes in the level of R&D work-force during the
«<
_! ge ' ' ' entire project period. as well as no big changes in the
eoooslg 1 ¥ ' ;
838S8.%.. 1x MIT 8S He EONS LG cost level.
2S BSS 6 ° ° . .
26h 6°83 $ 3 cs * Significant effects on productivity push this variable
¢ $ g $ to a peak of 2 units per man-month during 25-26 months.
MONTHS ——=—
* Project is completed by the end of month 35, 5 month lag
HENGH AMENSR = -PUCEAPPUC=F 0 SCOM=SFCI=F PROD=X behind the original schedule.
PPROD=Y CCOST=*
* Total effort needed reaches the minimun level of 700 man-
Fig.6 Schedule-First Policy
month, reflecting the rising of productivity.
50.00
20.000 30.000 40.000
10,000
218,
g8ss. AUR) hE BF SU) RES BS & BERWOCRTE FOF BIE | GRE DY Heveme J
Bas 2
nea! ' ' ! ¥
u ' 1 Mes, a
doneeee noednn nae | ;
a Perceived 1
1 ¥ 1 4
aa ' é|
$3888
g388, 0... auf wpsae 3
Bere
Noite j '
aaa Ff i
"
1 1 '
' wt '
Productivity \ t
;
ess 1
33 33 3
ge Bs
3B a8
a
1
o eS
8 88
ews
SS tok z fj ;
SEE 1 Ss 1
: ; 1 1
: « ; ; ;
7 ee 1 i 1
' at 1 Sst 1
esi ¢* 1 f
eo
$8 $85 nas
24 SO ° 2 2
ene 3 3 8 S
é 2 & 8 !
MONTHS
MEN=4 AMEN=R PJC=A PRIC=P SCOM=S FCn=F
PPROD=Y CCOST=* ‘
Fig.7 Fixed-manpower Policy
(5) IMMEDIATE SCHEDULE ADJUSTMENT POLICY +19,
Under this policy, schedules are inmediately adjusted to
correspond with changed forecast of project completion time,
This policy is much more similar to fixed-manpower: policy
through frequently and in-time adjustment of schedules. Such
policy usually submits to limiting level of funding or limited
pool of skillful R&D work-force with no strict limitation on
project completion date, such as research project at earlier
period, Alternatively, it may arise from the lack of the avai-
lability of additional R&D work-force to be assigned to the
project. In any one of above cases, once detected problems re~
sult in a later forecast completion date, the schedule is ad~
Jjusted to match the forecast, Under such a policy several cha-
racteristics can be found in the simulation results (see Fig.8)+
* Due to'no schedule pressure, there is no additional pres-
sure or motivation to change the rate of production of the
R&D work-forces.
* The organization size could be maintained at a stable
level, raising less than one person during the whole pro-
ject life
* A delay completion date with a big slippage of 6 months,
i.e, 20 percent over the original schedul
* As the penalty to the low productivity gains during the
project, with 726 man-month on the project (20 percent
over the ideal case of 600 man-wonth) end the highest cost
of 1,6 million yuan,
50.00
20.000 30.000 40.000
10.000
eoceo 120
S888. ee ee eee ee eee eee
2S 1 i ; f
a8ae!
ROH HD ee ee eH DD DH } <
' 1 ld tl at a at ee od
' i \ f
! ' Breeived ' KX
f - t ~
ose Bt Cumulative ac it
SSRs i Sk '
8828. ' p \
BERS
SS Sa i ; 1
s8a8"
' i ' f
f 1 j f
' ; \ f
1 j fi
'
22be i
$sss ‘
8888.
8B t 61
shoe?
'
\
ve i
. \
ggRe! vs decces
888 Soc Secor Stee
688! « ' yo:
i '
' ‘ Men ,
' ' ‘
\ ' '
' i \
eco t ' L
338 a nee
338 CI aia g
58 °
é&s 8
g
fod
8
MEN=4 AMEN=R PIC=A PRJC=P SCOM=S
aii Fcn=| i=
PPROD=Y¥ CCOST=* . r BaOREE
Fig.8 Intermediate Schedule Adjustment
294
+21,
BRIEF CONCLUT1ON
For the sake of policy choice, a table involving all the
simulation results indices of various policies mentioned above
is indicated in table 1.
Index Completion Cost Peaking Efforts
Date (million Workforce (man-
Policy (month) yuan) (men) month)
* Basic 33 1,56 26 709
* Accurate pro- 32 1.59 24 722
gress perception
* Schedule-first 31 2.02 31.5 713
* Fixed-manpower 35 1.54 20 700
(min. cost)
* Immediate sche- 36 1.60 20.5 726
dule adjust
* System Dynamics is an effective methodology, offering
a lot dynamic simulation data for policy analysis. Conce-
quently the policy will be selected according to the objec-
tives and strategy conducted by the environment.
* Each policy has its own characteristics and strategic
stress, as well as its advantages and shortcomings. The choice
of policy must be submitted to the main goal, i.e, the objec-
tives of the R&D plan, For instance, in our case, if the high-
est priority of planning is set on completion date, then the
policies of accurate progress perception and schedule-first
policy will be the best, Under the situation of limited work-
force, fixed-manpower policy and immediate schedule adjustment
7 222.
would be the best.
* Under particular major specified purpose a satisfactory
policy in most or ‘all dimensions can be searched though a
great deal simulation by system dynamics modeling. In this
case basic policy seems to be a satisfactory one in all di-
mensions.
* Concequently, system dynamics may serve as a multi-va-
Fiable decision making approach in R&D project planning, as
well as in the other cases.
Ref.
Edward B. Roberté, Nanagerial applications of
System Dynamics, 1978
APPENDIX
295
ODaze
nzr
DADD
4>22r
12230
RESEARCH AND DEVELOPMENT PROJECT MODEL
HRRRAERERE PROGRESS + RRRKEEREEE
=PCP. J+ (DT) (PPR, JRFPECK + JK)
MEN.K) (PPROD.K)
CF. K/ER
ER=1200
PCP--PERCEIVED CUMULATIVE PROGRESS (JOR UNIT)
EN ON JOB
“PERCEIVED PRODUCTIVITY ( JOR UNIT/MAN-MONTH )
PPJC--PERCEIVED PERCENTAGE OF JOB COMPLETED ¢ % >)
ER--EFFORT REQUIRED ( JOB UNITS >
=CRP. I+( DT) (PR. JK)
PR. KL=(AMEN.K) (PROD.K)
CRP--CUMULATIVE REAL PROGRESS (JOB UNITS)
PR--PROGRESS RATE (JOB UNIT / HONTH?
AMEN--AVERAGE MEN ON PROJECT
PECR.KL=(FER.K) (CRP.K-PCP.K)
FER. K=TABLE(TFER+PJC.K 90717062)
TFER=010907-57.8r1
PJC.K=CRP.K/ER
PECR-~PERCEIVED ERROR CORRECTION RATE (JOR UNE TNON THD
FER-~FRACTION OF ERROR RECOGNIZED ( % / MONTH
CRP--CUMULATIVE REAL PROGRESS
TFER--TABLE, FRACTION OF ERROR RECOGNIZEN ( % / MONTH )
PJC--PERCENTAGE OF JOB COMPLETED ( % )
PPROD. K=PPROD, J+ (DT) (CPPR, JK)
PPROD=PRODN
CPPR.KL=(FCPP.K) (PPROD.K)
FCPP.K=TABLE( TFCPPrRFSC,KrOr27,2)
TECPP= 25/4 23/62/+15/+08/0/-.05/-+07/—.11/-.21/- 44
CPPR--CHANGE IN PERCEIVED PRODUCTIVITY ( JOR UNITES/
MAN-MONTH/MONTH )
PRODN--PRODUCTIVITYs NORMAL (JOB unrTs/ MAN-MONTH )
Fee Ree TOME AP HANGE: IN PERCEIVED PRODUCYIVITY
iON’ ?
TFCPP--TABLEr FRACTIONAL CHANGE IN PERCEIVED FRODUC-
TIVITY ( % / MONTH >
4D4>pa0D
aD” aze onze DDD
ozr
22k,
RIK PRODUCTIVITY FOO
PROD. K=(PM.K) (PRODN) (NPF KD
PROM
PM. K=TARLE (TPMyRFSC KO 27.25)
TPM=.6/,65/.75/.85/1/1.3/1,5/1.6/1,45/1/.8
NPF-K=TABLE CTNPF e TIME .Kr 047296)
TNPF = 66/.7/.8/685/ 69/6 95/1/1.02/1.04/1.055/1.065/1.075/1,.082
NPF--NORM PERFORMANCE FACTOR (SKILL FROM PRACTICE OVER
TIME sDIMENSIONLESS)
CLIP--CLIP FUNCTION
FM~-PRODUCTIVITY MULTIPLIER
PRODN--PRODUCTIVITY » NORMAL (JOR UNIT/MAN-MONTH >
TPM--TABLE PRODUCTIVITY MULTIPLIER
RFSC.K=FCD.K/SCOM.K
RFSC--RATIO OF FORECAST TO SCHEDULER COMPLETION DATE
(RIMENSTONLESS >
FCD--FORECAST COMPLETION NATE (MONTHS)
SCOM--SCHEDULED COMPLETION DATE (MONTHS)
FCD.K=TIME.Kt+ITR.K
ITR-K=EBR,K/(PPROD.REMEN.K)
EBR.K=ER-PCP.K
ITR-~INDICATED TIME REMAINING (MONTHS)
EBR--EFFORT BELEIVED REMAINING (JOB UNITS)
SCOM.K=SCOM, J+ (DT) (1/0CS) (FCN, J-SCOM. J)
cD
6
DCS--DELAY IN CHANGING SCHEDULE (MONTHS)
ERRERKEREE MANPOWER HEEREERERE
MEN. K=MEN. J+ (DT) (MENCH. JK)
MEN=ER/ (DCOMI*¥PPROD>
DCOMI=30
MEN--MEN ON PROJECT
MENCH--MEN CHANGE RATE (MEN/MONTH)
DCMOI--DESIRED COMPLETION DATE INITIALLY ( MONTHS )
PPROD--PERCEIVED PRODUCTIVITY (JOB UNITS/MAN-MONTH)
MENCH.KL=FCHM.KEMEN,K
FCHM.K=TABLE ( TFCHMrRFSC.KrOr29 625)
TFCHM=- 665 /~ ¢4/~ + 2/~61/0/41/62/+4/665,
FCHM~-FRACTION CHANGE IN MANPOWER
TFCHM--TABLE» FRACTION CHANGE IN MANFOQWER ( % / MONTH )
AMEN. K=AMEN, J+ (DT) (1/DAMEN) (MEN. J-AMEN. J)
AMEN=MEN
DAMEN=1
DAMEN--DELAY IN AVERAGING MEN ON PROJECT (MONTHS)
296
EOE CHANGE IN COST FORE KE
CHEN. J4 (IT) (MENCH ¢ JK)
=WEPMMECHEN .
980 — YUANS
CHEN-~CUMULATIVE MEN ON PROJECT (MAN-MONTH)
CCOST--CUMULATIVE COST ( YUANS )
WEFMM--WAGE PLUS EXPENSES FER MAN-MONTH (YUANS)
SIMULATION SPECIFICATION HEEKEREEEE
ce i
FAO
a LENGTH. K=CLIF(Os50rPJC.Kr1) ¢
PLOT .MEN=Hr AMEN=R( 10950) /PIC=ArPPICHP (Or 1)/SCOM=S 7 FCI=F (25745)
X41 /PROD=X»PPROD=Y( «872. 1)/CCOST=*(0r 4E4
PRINT Oe eae PED PROUSEPRON. CHEN + AMEN» CCOST
SPEC DT=.5/PLTPER=1/FRTPER=3
RUN CONCISE MONEL
