The Industrial Base Analysis Model (IBAM)
Louis Edward Alfeld
Robert M. Sholtes
INTRODUCTION
IBAM is a production chain simulator. IBAM
simulates the flow of orders and products among the
many suppliers and subcontractors that combine to
produce an end item. The U.S. Department of Defense
has used IBAM to analyze the erosion of its critical
manufacturing base in an era of declining defense
budgets and to identify potential bottlenecks in the
event of renewed defense acquisition requirements.
IBAM can also provide support for analyzing industrial
competitiveness, planning regional economies, and
building “virtual” manufacturing organizations. The
model identifies key manufacturers, assesses labor
needs, forecasts technology impacts and prioritizes
policy options, all with little effort. The model benefits
users by pinpointing potential bottlenecks and
quantifying the relative costs and benefits of alternative
solutions.
A production chain may contain as many suppliers
as desired. Figure 1 provides an example of a
production chain developed by DoD to explore the
dynamics of helicopter manufacturing. Each of the
boxes in Figure 1 contains the name of a producer and a
product, joined together to form a production map.
Each box is actually a simulation model in itself.
During simulation, each box manufactures an end item
(or items) for shipment to the next fabricator up the
chain. Although every model on the map contains the
same generic structure, their parameter values all differ.
These different values define the individual
manufacturing processes and products that make up the
chain.
E21
f t t Zt a = t
Ea EA Pe Det Ce) 3) Ged
EIGIepeoeo | eee ees
Ss ———
ee] Co] Geer] a] Ean] Pe]
q i
t i]
[teen] [ettome] [Ensen] [Cem] een:
f t Ly
1 EA Ge
t
Woon) [Frencx | f
Figure 1: An Aggregate Helicopter Production Map
© 1995 Decision Dynamics, Inc.*
* Unpublished - All rights reserved under copyright laws of the United States
dU
ORDERBOOK
The program user must define a projected
orderbook profile for the end item. An example of an
orderbook profile is shown in Figure 2. The orderbook
profile may contain historical data as well as future
projections or assumptions concerning product demand.
More than one orderbook may be defined to reflect
multiple demand sources. Orderbooks may also be
defined for any or all of the suppliers and
subcontractors in the chain. At a minimum, IBAM
needs only a single, top-level orderbook to run a
simulation. If only one orderbook is defined, IBAM
will automatically set the orderbooks for the sub-tier
industries to establish an initial equilibrium. (During
simulation, actual orders for sub-tier suppliers will
depend upon demand from higher-tier producers.)
6
4
i
$BILLIONS
Figure 2: Orderbook profile for tactical missiles
Having defined a production chain and a top-level
product orderbook, IBAM then quantifies the impact of
changes in demand on all of the supplier production
capabilities. The model simulates the flow of orders
and shipments for each supplier in the production chain.
Figure 3 reproduces a representative IBAM output
graph based on torpedo production.
Coy T X
During simulation, production capacity across the
supplier chain adjusts to changes in orderbook
demands. In Figure 3, the initial drop in orders to the
prime contractor triggers a downsizing in
manufacturing capacity. In addition, every supplier in
the production chain adjusts their operations to the new
market conditions defined by the orderbook. The
hypothetical scenario shown in Figure 3 includes a
sudden new demand for torpedoes during 1999. This
jump in production demand causes suppliers to expand
capacity as each company acquires new capital, hires
and trains additional workers, and increases demands on
its own sub-tier suppliers.
MODEL ORGANIZATION
Figure 4 illustrates the primary feedback structure
of IBAM. Feedback acts to match shipments to orders
by altering production capacity. The model uses
incoming orders, average past demand and current
backlog to calculate a desired production rate. Capacity
is calculated using a Cobb-Douglas function that
depends upon both capital and labor. Actual production
is the lesser of either desired production or capacity.
Delays in perceptions of the average incoming order
rate coupled with varying delays in adjusting capital,
labor and materials generate the dynamic transients
typical of disequlibrium economic systems.
Orders |
Shipments
Desired
Production
Materials
C=}
NK.
shipments
cs
Dollars per month,
Figure 3: Capacity and shipments response to
orderbook
al
Figure 4: IBAM feedback structure
IBAM was specifically designed to operate in the
data-poor environment of industrial production.
Starting with whatever data is on hand, the user can
immediately build a production chain and run
simulations. As better data becomes available, the
model can be refined and extended to better satisfy user
needs.
IBAM contains an internal database with four-digit
SIC code data for 56 of the most common defense
manufacturing industries. By entering the appropriate
SIC code, IBAM will automatically supply the data for
an industry, interpolating values for capital and labor
based on.the ratio of user-defined shipments to total
SIC industry shipments and inserting the industry
average for shifts and utilization. The user may, of
course, override any of the program-supplied data
values with different values that reflect better or more
detailed data.
- The second input required of the user is the amount
of materials every manufacturer purchases from each of
its sub-tier suppliers. If this number is not known, the
user may “guestimate” the percentage distribution
instead. Given only this marginal information, IBAM
will provide a first approximation of the production
chain dynamics.
USER INTERFACE
Model users may build a production chain utilizing
a graphic toolbox that contains industry icons and
connector arrows. As each industry icon is placed on
the production map, it automatically creates an instance
of the generic simulation model. Double-clicking on
the icon then opens the data interface to that industry.
Figure 5 shows a sample data interface window.
Figure 5: Defining the Product
Data is organized on six tabbed displays. Figure 5
shows the display for the product variables. The left
side contains shipment data, including the SIC code
total for the industry. The right side shows the
parameters that define the production process.
IBAM is capable of using data at many different
levels of aggregation, from a single production line to
an entire industry. For some applications, the user may
wish to develop a detailed diagram of individual
production lines for critical components while
aggregating less critical material flows at much higher
levels.
“WHAT-IF?” SCENARIOS
IBAM “what-if?” simulations may be used to test
alternative parameter values and different intervention
actions. “What-if?” scenarios may also analyze
imagined products, proposed production chains and
even new process technologies about which there is
little or no historical production data. One model
application, for example, developed a hypothetical
production chain for the flat-panel display industry,
which has yet to develop. Making assumptions about
production parameters for each sub-tier supplier
allowed analysts to simulate flat-panel display
production under a variety of orderbook assumptions
that combined both defense and commercial demand.
The results showed relative rates of industry growth
under alternative DoD development and acquisition
scenarios.
IBAM includes a built-in “what-if?” technology
scenario generator, illustrated in Figure 6. The five
tabbed displays contain access to all model parameters
that reflect possible changes in product or process
technology.
Figure 6: Capital Technology
Figure 6 shows the capital technology display. The
user may alter any of five parameter values from their
baseline value, ramping the change over time, if
desired, to reflect the time required to implement a
specific change. Increasing capital productivity and
lowering new capital lifetime, for example, might
reflect the acquisition of a new computer technology.
Simulating these changes against a baseline
scenario will show the relative impact of the new
technology in response to an orderbook. Assessing the
cost of the technology upgrade against gains in capacity
or shortened delivery times allows industrial planners to
judge the relative value of alternative investment
Rr
options.
For example, Figure 7 plots model output for two
“what-if?” scenarios for helicopter production based on
a hypothetical jump in orderbook demand. In this
instance, only one helicopter, the RAH-66 Comanche,
was simulated by altering the parameter values used to
define the companies in the production chain.
=
| iY
£ Orderbook
i
&
g Scenario B
3S
i
a es
Figure 7: Relative production curves for two
simulations
In addition to its DoD aplications, the ability of
IBAM to create models for hypothetical industrial
chains affords users the ability to assess the dynamic
behavior of “virtual” manufacturing entities as well as
to study how to best nurture regional and local
manufacturing expansion. Economic planners may
“flight test” alternative industrial relationships to
determine which types of expansion best satisfy
regional and local employment, resource and
environmental goals.
REFERENCES:
Alfeld, Louis E. and Alan K. Graham, Introduction to
Urban Dynamics, MIT Press, Cambridge,
Massachusetts, 1976.
Clark, Rolf and Albert A. Pisani, “Defense Resource
Dynamics” Proceedings of the 1985 International
Conference of the System Dynamics Society,
Keystone, Colorado, 1985a, pp. 150-160.
Forrester, Jay W., Industrial Dynamics, MIT Press,
Cambridge, Massachusetts, 1959,
Forrester, Jay W., Principles of Systems, MIT Press,
Cambridge, Massachusetts, 1969.
Forrester, Jay W., “Growth Cycles,” De Economist,
vol. 125, November 4, 1977.
Forrester, Jay W., “An Alternative Approach to
Economic Policy--Macrobehavior from
Microstructure, “System Dynamics Group Working
Paper D-2876-1, MIT, Cambridge, Massachusetts,
1979.
Graham, Alan K., “The Long Wave,” Journal of
Business Planning and Forecasting, vol. 1, no. 5,
1982.
Graham, Alan K. and Peter M. Senge, “A Long-Wave
Hypothesis of Innovation,” Technological
Forecasting and Social Change, vol 17, 1980, pp.
283-311.
Hines, James H., “The Business Cycle and Money, An
Analysis of the Inventory Investment Hypothesis,”
Proceedings of the 1983 International System
Dynamics Conference, 1983, pp. 470-496.
Mass, Nathaniel J., Economic Cycles, MIT Press,
Cambridge, Massachusetts, 19XX.
Richardson, George P. and Alexander L. Pugh III,
Introduction to System Dynamics Modeling with
DYNAMO, MIT Press, Cambridge, Massachusetts,
1981.
Senge, Peter M. and Jay W. Forrester, “Tests for
Building Confidence in System Dynamics Models,”
TIMS Studies in Management Science, vol. 14, 1980,
pp. 209-228
Sterman, John D., “The Energy Transition and the
Economy: A System Dynamics Approach,” MIT
PhD dissertation, System Dynamics Group, Sloan
School of Management, MIT, Cambridge,
Massachusetts, 1981.
Sterman, John D., “A Dynamic Disequilibrium Model
of Energy-Economy Interactions,” International
Journal of Energy Sysiems, vol. 2, no. 3, 1982.
Sterman, John D., “Economic Vulnerability and the
Energy Transition, “Energy Systems and Policy,
vol. 7, no. 4, 1983.
Sterman, John D., “The Economic Long Wave: Theory
and Evidence,” Sloan School Working Paper WP-
1656-85, Massachusetts Institute of Technology,
Cambridge, Massachusetts, 1985.
Sterman, John D., “A Behavioral Model of the
Economic Long Wave,” Journal of Economic
Behavior and Organization, v. 6, 1985.
Sterman, John D., “An Integrated Theory of the
Economic Long Wave,” Futures, April 1985.
23
