Scheduling Methods For Ship Build Strategy Development:

Literature Search and Research Report

for the

Midterm Sealift Generic Build Strategy Task

The University of Michigan Transportation Research Institute,

Marine Systems Division

Transportation
Research Institute
 Technical Report J$b!I#lln Pie
1. RIporr NO. 2. Government Accession No. 9. RocipM's No.

fune B, 1997
4. Mk and S u b W 5, rpofl

Scheduling Methods for Ship Build Strategy 6. Performing Organlalon Coda

Development, Literature Search and Research Report 5670
8. Psdorming Orlynlzrtian Roped No.

-
ark H. Spickndl
9. Performing OgcmizaUon Nann m d Addmss
The University of Michigan
Transportation Research Institute 11. Contrltt or Gmnt No. I
2901 Baxter Road, Ann Arbor, Michigan 48 109-2150
12. Sporuoring Agency Name md Addmr
U.S.Navy through Designers & Planners, Inc. Technical
2120 Washington Blvd.
Suite 200 14. Sponsoring A w c y Codr
Arlington, VA 22204
15. Supplementary Notes

The purpose of this report is to identify existing and evolving scheduling approaches tlhat
could be used to help develop and evaluate build strategies for ships. A build strategy
embodies and communicates an overall plan for the production of a specific ship, or a
contracted series of ships, within a specific shipyard. The primary reasons for developing a
build strategy are:

(1) to ensure the particular shipbuilding program is feasible within the constraints imposed
by the shipyard, the customer, and the business environment
(2) to provide a framework for the coordination of work both within and across functions
in support of the program
(3) to provide a foundation for detailed/tactical operations planning that will generate
work packages, material orders, and shop-floor-level work sequences and schedules

1 17. b y W O ~ S I 18. DistributionStatement

build strategy, ship production planning,

production scheduling

19. ~ r m r t t ~lrrm.

Unclassified
y (al VIIS np0ti)

Unclassified
I
20. Swurlty Class~f.(of this page) 21. No. oi P a w

29 1
2 2 Prim
 Scheduling Methods For Ship Build Strategy Development:
1 9'396 (
Literature Search and Research Report
for the
Midterm Sealift Generic Build Strategy Task

Introduction

The purpose of this report is to identify existing and evolving scheduling approaches that
could be used to help develop and evaluate build strategies for ships. A build strategy embodies
and communicates an overall plan for the production of a specific ship, or a contracted series of
ships, within a specific shipyard.[lI1 The primary reasons for developing a build strategy are:

to assure the particular shipbuilding program is feasible within the constraints

imposed by the shipyard, the customer, and the business environment
. .to provide a framework for the coordination of work both within and across
functions in support of the program
(3) to provide a foundation for detailed/tactical operations planning that wi!l generate
work packages, material orders, and shop floor-level work sequences and
schedules

This investigation has been premised on the knowledge that world-class shipbuilding is
principally a group-technology-based fabrication and assembly process.[2] As such, the efficient
operation of a modem shipbuilding enterprise requires that planning and management focus
primarily on the aggregate use of resources across all contracts while also satisfying the
constraints of specific contracts. Correspondingly, an effective build strategy not only satisfies
the constraints of its associated contract, but also, when integrated into the overall operations
plan, allows that plan to make efficient use of a shipyard's resources across all anticipated
contracts. Such a planning and management approach is fundamentally different from the
project-oriented approach that is associated with one-off construction operations. Impoirtant
characteristics of modern shipbuilding and their implications for planning and scheduling are
described in this report.
Within a general manufacturing context, the processes associated with build strategy
development and integration are generically part of what is referred to as "medium-range
planning."[3][4] And the operations schedule developed as part of the medium-range planning
process, representing dl work of all contracts in aggregate, is typically called a "master
production schedule." This report describes both medium-range planning and prerequisite long-
range business planning as generally applied to manufacturing operations. Various scheduling
methods are described and evaluated relative to their applicability to build strategy development
and master production scheduling within a modern group-technology-based shipbuilding context.

' Numbers in brackets refer to references.

 Background: Characteristics of Modern Shipbuilding and Their Implications

General Production Approach

A ship will typically have the following general product characteristics:

extremely large
very complex
very high value
generally produced in low volume (one to ten per shipyard per year)
made to order
often at least semicustomized
a short required contract-to-delivery time relative to its work content

Many of these characteristics might lead one to initially conclude that a craft-based, one-
off construction approach is appropriate for producing ships. In fact, this is the production
approach that was prevalent throughout the world prior to WW 11. However, driven by the
urgent need for Allied ships during the war, Kaiser and Hann recognized that the principles of
group technology (GT) could be applied to shipbuilding.[5][6] This recognition was facilitated
by the evolution of welding technology and by Kaiser's and Hann's experiences within other
industries.
The GT-based approach dictates that ships can be progressively subdivided into
intermediate or interim products that can be classified into groups or ''families" based upon
commonality of production process. An interim product is any physical subdivision of a product
that objectifies a discrete set of work. The production process for each interim product group can
be rationalized to eliminate unnecessary tasks and to efficiently utilize focused resources,
sometimes resulting in a dedicated work cell or process lane. Each of these focused processes
can then be used repetitively on many similar interim products, resulting in significant
economies of scale at the interim product level. Also, using this approach, interim prodi~ctsof
different types can be produced in parallel and then assembled at later stages, greatly reducing
overall production cycle time.
World class shipbuilders have demonstrated that significant benefits can be realized fkom
GT even when concurrently building ships that vary somewhat in type, size, technology, and
detail, as there is still much similarity amongst their interim products throughout all stages of
production. Obviously, the more standard these ships are, the more focused and dedicated the
shipyard's production processes and tools can be, and the greater the potential gains. However,
GT requires that interim products be only similar enough to have common production p:rocesses.
In fact, the potential gains from increased standardization are very small relative to the gains
associated with initially just moving from a traditional construction-oriented approach to a GT
approach. The potential marginal gains from increased product and process standardization must
also be weighed against the resulting loss of business flexibilitylagility and product marketability
over time.
Implication: The medium-range planning process used must be appropriate for very
large, complex, high-value, low-volume, semicustom products that are made to order in a short
time pame using a group-technolou-basedproductionapproach.
Ship's Product Structure Characteristics
The process of building a ship can generally be represented by a hierarchical A-type
product structure, as shown in figure 1.[7] In such a process, parts are manufactured and
components are purchased, and these parts, components, and subsequent interim products are
progressively assembled to eventually form a single product. Such an assembly-based
production process has dependent demand, as each interim product depends on manufactured
parts, purchased components, andlor subassemblies fiom lower levels in the product structure.

Figure 1. Generic A-Type Product Structure.

A ship's product structure is not a pure hierarchical A-type structure, however, because
some many-to-many relationships exist between certain types of high-level ship interim
products, as shown in figure 2. Particularly, at the blocWassembly-to-zone level of a ship's
product structure, the erection of a single block or major assembly can play a part in the creation
of several on-board outfitting zones, and a single on-board outfitting zone is likely to have been
created by the erection of more than one block or assembly. Similarly, at the zone-to-ship
system level, on-board outfitting work in a single zone can be prerequisite to the completion and
testing of several ship systems, and a single system can be present in several zones.

Figure 2. Many-To-Many Relationships In Upper Levels of a Ship's Product Structure.

Implication: The medium-range planning process used must account for dependent
demand. Also, because the medium-range planning process uses the interim products in the
upper levels of a ship's product structure to derive the ship's work, it must be able to accountfor
the many-to-many relationships that exist there.
Operations Management Orientation
Because there are typically many interim products of several ships continually competing
for the finite capacity of a shipyard's resources, planning for and managing such an opecation
solely fiom the ship/contract/project perspective, as would be done within a traditional
construction context, will result in competing, uncoordinated, and erratic demand on individual
production resources. This, in turn, will result in tremendous inefficiency, delays, disruption
costs, and work-in-process (WIP) inventory costs. Therefore, the primary operations planning
and management task is to optimize process/resource utilization for interim products across all
contracts in aggregate.
However, because of the high value of each ship, ownershuyers with a project focus have
and exercise significant power over shipyards. Also, because shipyard cash flow is usually tied
to key project-specific milestones, significant attention should be paid to the potential fi~wmcial
implications of project-specific planning decisions. So managing such an operation solely from
a aggregate perspective could result in late ship deliveries and/or lost financial opportuiities.
Implications: The medium-rangeplanning process used must be able to simulta~rteously
address aggregate resource and material issues, andproject/ship-specifzc dependent demand,
work criticality, conpact milestones, and delivery schedule constraints.

Systern Variability
VariabilityIn Process Performance - Significant levels of performance variability will
occur in shipyards that have not fully or effectively established GT-based interim product
families and associated processes, or implemented SQCISPC procedures throughout operations
to maintain and improve process control. This performance variability and its accumulation
through a ship's many stages of interim product concatenation can have a major impact on
overall production system predictability and performance, and associated cost and scheclule
risk. [8]
VariabilityIn Demand - Demand for a shipyard's products and capacity can be clifficult
to predict. For planning purposes, demand can be dealt with either deterministically or
stochastically depending on the circumstances. If there is no uncertainty about demand for the
shipyard's capacity over the medium-range planning time horizon, the planning problem is
deterministic. This would likely be the case if a shipyard includes only current work and highly
probable proposed work in its medium-range planning process.
If the medium-range planning process includes ships that have only been forecast,
stochastic representation of at least some of the forecast demand may be necessary depelnding the
length of the planning horizon, the reliability of forecasts, the overall size of the product's market
and the level of competition in that market. Following are two examples of using detenministic
versus stochastic data in planning.

Example 1 : A shipyard has the capacity to competitively produce four ships of a

particular type per year, and knows that there are three competitors that can do the
same. The medium-range forecast identifies total demand ranging fiom 10 to 15
of these ships in each of the next two years, with 11 ships per year being most
. likely. In this case, the stochastic nature of this forecast is very important for this
shipyard's planning. Conversely, if the medium-range forecast identifies an
expected demand of 15 to 20 of these ships in each of the next two years, the
 shipyard can probably use a deterministic planning approach that assumes a
steady demand of four ships per year.

Example 2: A shipyard is competing for two independent contracts for which the
start dates are known. The probability of winning contract A is determined to be
95%, while the probability of winning contract B is determined to be 50%. In this
case, planning should probably treat contract A deterministically and contract B
stochastically.

Implications: The medium-range planning process should be able to accountfor process

variability through stochastic representation of associated variables, and also should be able to
employ a stochastic representation of demand fi necessary. When stochastic representations of
process performance andlor demand are used, the medium-range planning process should
generate probabilistic schedules, resource utilization plans, and cost estimates, , both for risk
assessment and to support the development and inclusion of risk mitigation strategies as part of
the overall build strategy.l The goal of stochastic planning is to generate a plan that has a
probability of occurrence that is acceptable @om the standpoint of satisaing both the customers'
needs and the company's strategv and business plans; that is, that represents acceptable levels of
business risk for the company and its customers.

Degree of Individual Product Uniqueness

Some commercial ship market segments, like the bulk carrier segment, are primarily
commodity-like, while other market segments, like the cruise ships segment, are made up almost
entirely of semicustom or custom ships. Most commercial ship market segments, however, have
a low-end, commodity-like subsegment and a higher-end, semicustom or custom subsegment.
Within a commodity segment or subsegment, ships have essentially equivalent features
and capabilities, and price and contract-to-delivery cycle time are the only significant marketing
characteristics. In a commodity market, producers who want business are forced to accept the
price and cycle time performance set by the pricelperformance leaders with available production
capacitya3Because of the resulting price and performance pressure, profit per unit of production
is usually very small. So commodity producers attempt to earn acceptable profit levels and
improve performance by increasing production capacity and volume, that is, market share. This,
in turn, adds even more price and performance pressure to this market. Poorer performing
producers can not profitably compete in commodity markets unless demand is so high that the
capacity of all the better performing producers in that market is hlly utilized.
Within higher-end market segment or subsegments, customers are looking for added
value along a number of additional product dimensions, including:

improved financing
reduced concept-to-delivery cycle times
higher quality
improved functional and qualitative features

bbSirnulation"in this instance refers to Monte Carlo-type simulation.

' Price and cost in this context include the effects of any relevant subsidies.
 improved technology
lower operations costs, including costs associated with normal operations,
reliability, maintenance, and overhaul
increased life cycle support

Identifying and meeting customer needs in higher-end market segments requires ,a great
deal more work in marketing, product development, and planning. But the potential rewards are
great as the shipbuilder can then set its prices according to each unique ship's perceived
differentiated value. [9]
Implications: Because of the narrow profit margins and strict delivery
requirements associated with commodity ships, scheduling and cost estimating in
support of commodity shipbuilding must be extremely accurate. Also, because of the
commonality of such ships, much of the up-fiont design andplanning work is complete
at contract award and the vast majority of the contract-to-delivery time and cost is
associated with material control and production. So medium-range planning will be
focused on rhese aspects of operations with fairly complete design information and
planning standards being available. For semicustom and custom ships, a signijkantly
larger portion of contract-to-delivery time and cost will be associated with product
development, design, engineering, material procurement, andplanning. Therefore, the
medium-range planningprocess for such ships must be much broader in scope and
focused on overall coordination of work associated with all importantproduct
dimensions. This planning process must also be able to deal with incomplete and
evolving information.

Similar Industries
Other moderate volume producers of complex, high-value, assembled products include
the aerospace, heavy-equipment, large-machinery, and large-machine-tool industries. Ships will
typically have a higher level of work content and complexity and a lower level of product and
interim product standardization than the products of these other industries, with some specific
exceptions. Shipbuilding has also sometimes been compared to large-scale industrial plant and
building construction. But while some large-scale construction enterprises utilize a GT-based
approach for some fabrication and subassembly work at dedicated off-site facilities, industrial
plant and building production work is still planned and managed primarily as project-oriented,
on-site construction.
Implications: Shipbuilding has much in common with other manufacturing industries,
particularly when viewed in a GT-based context. Therefore, much can probably be learned and
appliedfiom the existing body of knowledge in general manufacturing, operations management,
and production planning.
 Description of Long-Range and Medium-Range Manufacturing Planning Processes

Following is an overview of how manufacturing companies typically develop long-range

and medium-range operations plans and schedules. Figure 3 shows the outputs of these planning
processes and their relationships. This overview summarizes these processes as they are
presented in current operations management and production planning literature.[lO][l 11

7)
I ong- ange annlng
Information and Forecasts LONG-RANGE STRATEGIC PLANNING

Long-Range Resource
Finacial Plan +-+ Requirements
Plan

Overall Capadty
and Cost Check

\ I New Product Overall Schedule Check I

Aggregate
Production Master

Rough-Cut
Capacity Plan

IDETAILED PLANNING J
Figure 3. Typical Long-Range and Medium-Range Manufacturing Planning Process.
The Long-Range I Strategic Planning Process
A business will typically carry out long-range planning annually, or when the business
environment changes significantly, using the current status and long-range projections for each
of the following:

general macroeconomic climate

market demand for product(s)
money markets
human resource markets
capital goods markets (hard technologies)
subcontractor markets
raw material and purchased component markets
political environment
the company's capabilities and backlog
potential competitors' expected backlogs, capabilities and actions

Long-range planning can have a ten-year time horizon with planning considered in one-
year increments or "time buckets," although this can vary significantly by product type. For
instance, long-range planning in the computer industry usually has a time horizon of three years
or less because of the pace of associated technological change. Long-range planning outputs
typically include the following items:

a business strategy that articulates product lines, organizational structure,

quality goals, pricing goals, delivery and service level goals, market-share
and production-volume goals, profitability goals, etc. over the planning
horizon
business policies andprocedures that, within the context of the overall
business strategy, represent process and product standards for marketing,
design, operations, finance, purchasing, logistics, human resources, and
cross-functional activities
a resource-requirements-and-capacity plan that identifies the capacity and
associated facilities, equipment, personnel, and subcontractor and supplier
relationships that will be required to meet the long-range goals identified in
the business strategy
aproduct research-and-development plan to meet the perceived needs of the
market as articulated in the business strategy
afinancialplan that describes the approaches to the money markets and
share holders that will be used to support the business strategy, the long-
range resource-requirements-and-capacityplan, and the long-range product-
development plan
The Medium-Range Planning Process
Medium-range plans
. - are typically revised monthly or quarterly, or as the business
situation changes, using the long-range plans as a point i f dep&ture. Medium-range planning
typically has a time horizon of 12 to 60 months, and time buckets of anywhere fiom a day to a
month, depending on product type and complexity, typical production cycle time, demand
forecast variability, and the associated level of risk the company is willing to accept. Medium-
range glans account for all certain work and can also include proposed and forecast work;. The
medium-range planning process and associated build strategies will be constrained by thle
shipyard's business strategy, long-range business plans, operations policies and procedures,
capabilities and capacities, and anticipated concurrent work, customer requirements, the cost of
capital, ships' characteristics, and the anticipated availability of material, components, and
associated information. Medium-range planning typically results in the creation of the following
outputs.

An aggregate production plan (APP) specifies total projected aggregate

demand and output either per time bucket or cumulatively over time. It is
developed to identify the overall resource and inventory utilization strategy
that is most likely to minimize total operations costs over the planning
period while satisfying the serviceldelivery, quality, and overall production
volume goals defined in the long-range plan. Such a strategy will seek to
minimize cost over the planning horizon by varying the relative use of in-
house production capacity, inventory/backlogging, and subcontracting to
satisfy overall demand. Total aggregate demand and output are expressed
by a surrogate attribute common to all products, like direct labor hours, units
of production, or tons. At one extreme, the APP could reflect a "chase"
strategy where in-house capacityloutput is varied continually over time to
exactly match expected aggregate demand. At the other extreme, the APP
could reflect a constant capacity strategy where incremental demand greater
than this capacity would be handled with overtime or subcontracting, and/or
worked earlier at times when demand is less than capacity and then stored in
inventory. Nearly all real-life APP strategies fall somewhere between these
two extremes. Reference [lo] presents an excellent APP example. Various
methods are used to develop APPs including manual iteration called "cut
and try," mathematical modeling, computer-based heuristic modeling, and
simulatioll of masler production schedules (see below).

A master production schedule (MPS) specifies dates and quantities of

production for each specific product, or for each of a number of common
upper level interim products (called "end items" or "planning units") for
complex products, within the constraints of the APP. The MPS serves as the:
basis for ensuring and reserving adequate capacity of critical production
resources over time for all work to be accomplished (see "rough-cut capacity
plan" below), generating detailed ("shop floor") production plans and
schedules, and establishing a material procurement schedule. Other inputs
 to the MPS include the structured bills of material for all products down to
the planning unit interim product level (for the identification of planning
units and their associated dependencies), and the process standards
documented in the company's policies and procedures (used for the
identification of work breakdown structures, process durations, and resource
requirements for each type of product or planning unit).

r A rough-cut capacityplan (RCP)identifies the expected usage of critical

resources over time for all products or planning units in support of the MPS.
The RCP is compared to the long-range resource requirements and capacity
plan to assure that adequate resources will be available within the medium-
range time horizon. If it is determined that the long-range resource
requirements and capacity plan does not support the critical resource needs
dictated by the MPS, management can modify the long-range resource-
requirements-and-capacity plan, the APP, and/or the MPS. Note that any
changes to the long-range resource-requirements-and-capacity plan will
likely require changes in the other long-range and medium-range plans.

Planning For A New Product

When a new product is proposed for production, the management team must first
determine if the product fits within its business strategy, existing policies and procedures, and its
other long-range plans. If the product obviously does not fit, management must either drop the
product fiom consideration or modify the organization's long-range plans.
If the product clears the initial strategic hurdle, a new iteration of the APP is generated
including the new product and any additional copies that customers will desire. In doing this, the
overall capacity reservations over time for all previously committed/ contracted products can be
changed as long as their critical schedule milestones are not altered. The regenerated APP
should provide a good overall indication of:

r the new product's overall production duration

r the overall capacity, overtime, subcontracting and inventory requirements
that will most economically satisfy aggregate demand for all products,
including the new product, through the medium-range planning horizon
r whether the initial new product can be produced within the required time
constraint
r whether follow copies of the new product can be delivered at the desired rate
If there are overall capacity problems identified at this point, the company may not be
able to produce the product. Or the long-range plans might be revisited to see if changes might
be made to profitably accommodate the product.
If the product makes it past the APP hurdle, the MPS and RCP are generated to include
the new product. The MPS must support the APP, and the RCP must not exceed the key
resource limitations designated in the long-range resource-requirements-and capacity-plan. The
MPS will identify times and quantities of production for each planning unit while also
accounting for the capacity constraints of specific resource and intraproduct dependent demand.
The RCP will identify the utilization levels of key resources over time across all products. The
new product's production schedule and critical-resource-utilization plans can be extracted from
the MPS and the RCP, respectively, at the planning unit level. Because the APP, M P S , and RCP
are so interrelated, they evolve in an integrated and interactive way. Once again, the overall
objective of'this planning process is to minimize business risk and improve the probability of
meeting the goals of the shipyard and its customers.
 Literature Search Results

The literature search was conducted to identify medium-range scheduling methods and
tools that are being used today in manufacturing environments similar to shipbuilding, and also
methods that are being researched and developed for potential future application in these
environments. Following are brief descriptions of the general methods that have been identified.
It is beyond the scope of this report to describe these methods in detail and provide working
examples, so the reader is referred to the references for more detailed descriptions and
information.

Search-Based Heuristics
Some simple heuristics have been shown to find optimum product sequencing solutions
for some very simple independent-demandjob-shop and flow-shop production systems (:like n
jobs of known duration and one machine, and n jobs of known duration and two machines in
sequence). Examples of these heuristics include first come first serve, earliest due date first,
shortest processing time first, longest processing time first, least slack first, etc. These types of
heuristics are not adequate by themselves for finding useful solutions for more complex medium-
range planning problems. However, such heuristics can sometimes be usehl for determining
near-optimum work sequences and schedules when used in combination andlor with other
scheduling approaches. Search-based heuristic methods, like branch-and-bound and branch-and-
cut, are methods for utilizing rules to narrow the solution space of a combinatorial optirriization
problem so as to efficiently obtaining a "good" solution. There is at least one know application
of this approach to medium-range planning and scheduling of ship production in

Mathematical Modeling
Mathematical modeling generally involves the optimization of one or more objectives
that are represented mathematically (e.g., "maximize profit") subject to a set of constraiints that
are also represented mathematically (e.g., panel line usage <= panel line capacity of 3000
hourslyear). Common approaches of mathematical modeling are linear, integer and nonllinear
programming, goal programming for multiple objectives, and dynamic programming for such
problems solved over time. These techniques were developed in the 1950s, and are conunonly
used for capacity planning and scheduling in continuous processing industries like oil refining,
steel, chemicals, and paper manufacturing, and in assembly, transportation, and service
industries. The outputs of these methods will typically include sensitivity analyses of input
variables for risk assessment. Although mathematical models are generally intuitive, a 'high level
of competence is required for their development and use in real-world situations.[l9][20][21][22]

Network Scheduling
Network scheduling generally represents the critical path method (CPM), program
- -

evaluation and review technique (PERT), graphic evaluation and review technique (GE:RT), and
probabilistic network evaluation technique (PNET). Network scheduling is typically project
focused, represents work via each project's networked work breakdown structure (WBS) tasks,
backward and forward schedules the project tasks, and identifies the project's critical tasks and
path, and the slack or float on other tasks. Most network scheduling software packages support
the identification of resources and material to tasks, and thus allow the generation of resource
and material constrained project schedules, as well as utilization plans for the resources and
material themselves. These packages also generally support the identification of costs to .
resources so that cost estimates can be generated integrally with schedules and resource
utilization plans. Some approaches support the simultaneous scheduling of multiple projects
sharing a common pool of resources, and some support stochastic representation of task
durations and costs, thus allowing probabilistic schedule simulation for the assessment of
schedule and cost risk. Network scheduling is probably the scheduling approach that is most
familiar to U.S. shipbuilders, although it is not clear that these shipyards utilize all of this
method's capabilities.[23] [24][25][26][27][28] [29][30][3 1][32][33]

BOM-Based Backward Scheduling With Standard Lead Times

This approach generally represents the material requirements planning (MRP)
methodology in which work is represented by hierarchically structured bills of material for
dependent-demand assembled products, together with associated standard production lead times
for each interim product. Interim products are backward scheduled down the legs of the product
structures from their final products' or end items' due dates using the standard lead times. This
methodology has continually been refined ("closed-loop" MRP, MRP 11, ERP) to account for
finite capacity and nonhierarchical substructures (many-to-many relationships) in the BOM, and
to more efficiently schedule all resources relative to aggregated product demand. This approach
was developed by IBM in the mid-1 960s along with hierarchical database technology, and is
information system intensive. Some U.S. shipyards have attempted to use an MRP-type
approach for detailed-level planning and scheduling.[341[351[361[3 71

Synchronous Manufacturing Scheduling

This approach is based on what is commonly known as the theory of constraints. It
focuses on identifying the bottleneck resource in the process, "exploiting" that constraint to the
greatest extent possible (finding ways to maximize its throughput in support of demand), and
then "subordinating" all other processes to the "drumbeat" of the bottleneck. Work is sequenced
at the bottleneck to support final demand. Then dates for release of material from the bottom of
the product structure into the production process are backward scheduled from the bottleneck's
requirements using the length of "time buffers" between release and the bottleneck, at the
bottleneck itself, and between the bottleneck and final product delivery. These time buffers will
include setup time, process time, queue time, wait time, and idle time. Capacity is accounted for
explicitly at each step in the scheduling process. Strict heuristics establish priorities for job
sequencing at the bottleneck (earliest due date first) and for increasing throughput at resources
that are not bottlenecks but that are temporarily constrained: (1) batch products if possible to
reduce setups (increasing WIP), (2) use overtime specifically for critical jobs, (3) off-load critical
jobs to other resources, (4) delay final product delivery. This approach has tended to be more of
a philosophy than a formal methodology because of the lack of objective documentation of the
methodology and its implementation in industry. Most U.S. shipyards are familiar with, and
agree with, the hdamental ideas articulated by the theory of constraints. [381[39][40][41]
 -
Discrete Events Simulation
This approach models the production process as a complex set of interrelated discrete
events that occur through time. Within a simulation model, production resources, material
handling entities, praductlinterim product entities, and information entities are modeled and their
logical relationships defined. Each entity includes the definition of specific attributes that will
impact system performance. Production is simulated by the product entities actually going
through the discrete events of production in time simulated by the computer's clock. Many
entity characteristics can be represented either deterministically or stochastically, and
probabilistic results can be obtained. Discrete events simulation tools do not normally
incorporate work sequencing and scheduling functions, so simulation by itself cannot be
considered a scheduling approach. Therefore, for scheduling purposes, it must be combined with
other rnethodologies (see "Agent-Based Systems" below), and really serves as both a tool for
validating a given schedule within the production system, and potentially as a graphic user
interface, A significant level of expertise is required to build, run, and interpret realistic
simulation models, although simulation systems are evolving to include more intuitive interfaces
and standard libraries of common servers and other entities. Many U.S. shipyards are familiar
with discrete events simulation primarily from a process design perspective rather than al
planning and scheduling perspective. However, at least one U.S. shipyard has successfully
incorporated discrete events simulation into detailed shop-floor-level scheduling of its panel
line.[42][43][44]

Expert Systems
An expert system is a set of hierarchical heuristics or rules that infer the actions and
interactions of objects in a specific domain with the objective of making a particular decision in
that domain. Expert systems are useful for solving problems in domains where qualitative issues
must be addressed. The logic within an expert system is structured and is often based on the
"rules of thumb" that an "expert" has derived through experience to get "good" workable
solutions to a specific problem. The concept of expert systems is known to some U.S. shipyard.
[451[461[471[481

Agent-Based Systems
An agent-based system is also a heuristic system. However, rule sets are embodied in the
objects within the system that interact to attempt to accomplish certain goals. Each object's rules
dictate how that object will react to encounters with, and the actions of, other agents andl the
environment. The rules that govern an agent's behavior may also change according to its
interactions. Current agent-based scheduling systems are implemented within a discrete-events-
simulation framework. This approach to heuristic-based system modeling is much less
structured than an expert system, and is thus more flexible. In many respects, this approach
resembles real-life domains where people, objects, and information interact, negotiate, rnake
decisions, and solve problems simultaneously according to individual and global needs and the
conditions of the domain at that time. Agent-based applications to planning and scheduling are
relatively new to all industries, but there are some relevant references available.[49][50][5 11
Systems Dynamics Simulation
Also known as numerical simulation, systems dynamics is an approach that attempts to
identify and characterize mathematically all of the interactions between objects and activities in a
particular system, and then identify how certain specific changes to the system will propagate
and impact the individual elements of the system and the system as a whole. Its focus is on the
prediction of impacts fiom system changes, or contingency planning, not necessarily initial
planning. The objective is similar to agent-based systems, except in this case the system's
behavior is described mathematically rather than heuristically. Decision Dynamics, Inc. has
done some work in applying systems dynamics to predicting outcomes of changes in ship
production scenarios. Some of the shipyards in this project have beta tested DDI's
sohare.[52][53]
 Evaluation of Medium-Range Master Scheduling Methods

Following is a general evaluation of the applicability of the scheduling methods identifie:d above
to the ship production domain. This evaluation criteria were developed by UMTRI-MSI) based
on the perceived needs of U.S. shipyards and the Navy, and were reviewed by the BSM team and
modified as necessary. UMTRI-MSD then evaluated the various methods versus these criteria.
Following the descriptions of the criteria, an Evaluation Results Matrix is provided along with an
Evaluation Summary and Recommendations.

Evaluation Criteria
Easy to Understand and Use
Identifies whether a typical production planner within a US shipyard would find a
particular method easy to comprehend and associated scheduling models easy to build.
"Yes," "Somewhat," "No."

Familiar to US Shipbuilders
Measured in degree. "Yes," "Somewhat," "No."

Addresses Finite Capacity I Resource-Constraints

Simply identifies whether the method always assumes infinite capacity, or whether it
supports finite capacity or resource-constrained scheduling, including resource 1e:veling.
"Yes" or "NO,"

Allows for Different Dependency Types Between Tasks or Interim Products

Sometimes the performance of a dependent task is not constrained only by the fiinish of
other tasks (finish-to-start constraint, FS). Sometimes start-to-start (SS), start-to-finish
(SF), and finish-to-finish (FF) constraints more realistically represent the
interdependencies between some tasks. For example, the testing of a particular ship
system may be able to start afier the assembly of its first outfit unit, but cannot be
finished until some period of time (lag) after its last outfit unit is erected and related on-
board outfitting work in that zone is finished. This identifies whether a particular method
allows the use of these different types of dependencies and their associated lag times.
"Yes" or "No."

Allows the Use of Heuristics for Task Prioritization

Sometimes there may be other factors that should influence scheduling that can be
expressed by heuristics. For example, if two flat blocks are scheduled to start assembly at
exactly the same time but there is only one flat block assembly bay available, a heuristic
might state that the block with the least total project slack has priority. Another example:
one might want particular tasks to start as late as possible without influencing th'e
completion date of the project to support a just-in-time operations policy. This rneasure
identifies whether a particular method allows for the use of user-defined heuristics
. beyond the intraproduct dependencies described above. "Yes" or "No."
Performs Forward and Backward Scheduling to Identify Critical Paths and Tasks
Forward and backward scheduling establishes time windows in which each specific task
can be scheduled without impacting overall projectJship schedules. This capability can
provide information about task float or slack, critical tasks, and schedule flexibility that
management can use to make tradeoffs relative to resource utilization, inventories,
material procurement, cash flow, risk, etc. "Yes" or "No."

Provides Aggregate Resource Management Perspective

Planners and managers are able to review the schedules and utilization of key resources
(human and capital) across all ship contracts, and identify and manipulate tasks within
any project and resources that are constrained or critical in order to facilitate improved
aggregate resource management and eficiency. "Yes" or "No."

Provides Project.Contract Management Perspective

Planners and managers are able to review work schedules for individual projectstship
contracts, identify and manipulate those tasks that are critical, and keep track of key
project dates to facilitate improved projectJcontract management. "Yes" or "No."

Provides Material Management Perspective

Planners and managers are able to review material schedules across all ship contracts and
identify if and where there are potential material procurement problems for improving
material management. "Yes" or "No."

Supports Easy and Fast Scenario Analyses ("What if.. .")

Changes to inputs and constraints are easy to make, and alternative schedule scenarios
can be created, analyzed, and compared quickly. "Yes" or b'No."

Provides Optimization Capability

The approach has a means of iteratively revising and improving the schedule relative to
specified objectives. Such capability can be implemented through the use of such things
as genetic algorithms and search-based heuristic methods. "Yes" or "No."

Determines Realistic Impacts of Potential Work-In-Progress Changes

The impact of plan changes during the production process is typically nonlinear. For
example, the addition of work during production will not simply result in the addition of
that work's labor hours and time to the existing plan because of disruption and the
replanning required to make the best use of resources within the changed constraints.
While analyzing the impact of in-process changes may not be part of medium-range
planning for some, others may want to cany out realistic contingency planning at this
stage. "Yes" or "No."

Provides Risk Analysis Capability

Supports schedule risk assessment through the determination of probabilistic outcomes
based on stochastic inputs andlor sensitivity analysis of inputs. "Yes" or "No."
a Commercial Software Available
Identifies whether there is commercially available software that supports a particudar
method. "Yes" or "No."

a Software Easily Integrates With Other Planning Tasks

If software is commercially available to support a particular method, it can be easily
integrated with front-end processes and their applications, such as a structured BOM and
process definition standards, other medium-term processes and their applications, such as
task duration calculation, APP, RCP, and RCP versus Long-Range-Resource-
Requirements-and-Capacity Plan comparison, and follow tasks such as report generation,
long-lead-time material procurement, and detailed shop-floor planning and scheduling.
"Yes,""Some!what," "No."

Evaluation Results Matrix

U Vi
Y

-
V1
Vi y
C
2
2
U
aJ
m
i,

w d LO
g .z
c
sm m
.E
n.
V1
L .c:
-,
c:
E

"
4
c:
3
m
X!
w
U
a 0
V1

Evaluation Criteria - U

Medium-Range Scheduling Approaches

Heuristics S S Y Y Y N Y Y N N Y N N N M
Mathematical Modeling N S Y Y Y N Y Y Y Y Y Y Y N M
Network Scheduling Y Y Y Y Y Y Y Y Y Y N N Y Y S
BOM-Based Scheduling w/ Std. Lead Times Y Y Y N Y N Y N Y Y N N N Y Y
Synchronous Manufacturing Scheduling S S Y Y N N Y Y Y N N N N Y 1
Discrete Events Simulation S S Y Y Y N Y N Y N N Y Y N S
Expert Systems S S Y Y Y N Y Y Y Y N N N N S
Agent-Based Systems ? N Y Y Y ? Y Y Y Y Y Y ? Y ?
Systems Dynamics Simulation S S Y N Y N Y Y N Y N Y N N S

Evaluation Summary and Recommendations

In the short term, all the capabilities of the network scheduling approach should be
utilized to accomplish the master production scheduling process necessary for build strategy
development. This approach is the most familiar to U.S. shipbuilders and supports much of the
req'uired functionality. A full-featured network scheduling application like Primavera sh~ouldbe
used. In addition, associated research and development should be carried out in the areas of
 schedule optimization, realistic contingency analysis (perhaps through integration with systems
dynamics simulation), realistic risk analysis, and integration with other planning functions.
Also in the short term, the heuristic scheduling approach described by Lee et. al. [13]
should be implemented in a generalized form. Research and development in the areas of realistic
contingency analysis, risk analysis, and integration with other planning functions should also be
carried out in support of this approach. Once developed, it could be benchmarked against the
network scheduling methodology for ease of use and usefulness.
In the medium term, the applicability of an agent-based approach to master production
scheduling should be investigated because of its potential for accurately emulating real-world
planning operations and providing inherently realistic and useful results. This approach would
be implemented with a discrete-event-simulation interface.
In spite of its significant capabilities, mathematical modeling is not likely to be
implemented within some shipyards because developing, implementing, and understanding
useful models is difficult.
Although BOM-based scheduling with standard lead times was specifically developed to
deal with dependent-demand assembled products in aggregate, it typically assumes that only
finish-to-start constraints exist between interim products in different levels of the product
structure. This method also sometimes has difliculty with many-to-many relationships within
the product structure. Only backward schedules are produced by this method and thus it does not
provide information about project-specific critical tasks and paths. The method also typically
assumes that lead times are deterministic, thus preventing risk analysis. While soAware
developers have continued to refine this approach to attempt to address some of these problems,
it is probably not the best approach for master production scheduling. However, it might still
represent the best way to "explode" the master production schedule down into more detailed
shop-floor plans and schedules, as identified by Neurnann.[32]
Synchronous manufacturing scheduling is too undefined in the literature at this point to
determine its usefulness for master production scheduling. Goldratt makes some broad
generalizations in outlining this scheduling approach that practitioners continue to question. For
example, he suggests that the first strategy for increasing the capacity of a temporarily
constrained resource is always to increase batch size, thus increasing WIP, when other strategies,
such as temporarily adding overtime, might have more economic merit in some circumstances.
There is software available that purports to use this methodology for shop-floor level scheduling.
Perhaps this software could be obtained and exercised to ascertain its potential validity and
usefulness for medium-range planning.
Discrete-events simulation by itself has no inherent scheduling capability, so it must be
interfaced with some other sequencing and dispatching tools. It is also not capable of identifying
critical paths and tasks and thus providing a project management perspective. Its best potential
use with respect to master production scheduling might be to serve as a validation tool and
interface for an agent-based scheduling system.
An expert system is made up of a structured hierarchical set of heuristics that have been
found to produce "good" and usefil solutions for a particular domain. If a particular expert
system is applied to a domain for which it was not intended, it will not produce useful solutions.
Because U.S. shipbuilders are likely to produce a variety of products, and because yards vary
considerably in their capabilities, it is not likely that a single expert system can be developed as
the master production scheduling tool for all potential shipbuilding situations.
 Systems-dynamics simulation is best used for analyzing the impact of changes to a
system based on empirically developed relationships between system entities. In this regard, it
might be most useful for scenario analysis and optimization when used in conjunction with other
scheduling approaches, rather than for initial system modeling.
 References

Book titles are underlined, technical report, journal and proceedings titles are italicized, (and
paper titles are in quotes.

[I] Lamb, T., Build Strategv Development, National Shipbuilding Research Program, WSRP
0406,1994.

[2] Storch, R. L., Hammon, C. P., Bunch, H. M., and Moore, R. C., Ship Production, Ijecond
Edition, Cornell Maritime Press, Centreville, MD,1996, pp. 45-53.

[3] Chase, R. and Aquilano, N., Production and Operations Management - A Life Cycle
Approach, Sixth Edition, Richard D. Irwin, Inc., Homewood, IL, 1992, pp. 608-611.

[4] Narasimhan, S., McLeavey, D., Billington, P., Production Planning and Inventory Control,
Second Edition, Prentice Hall Inc., Englewood Cliffs, NJ, 1995, pp. 251-253.

[5] Chirillo, L. and Chirillo, R., "The History of Modem Shipbuilding Methods: The U.S.-
Japan Interchange," Journal of Ship Production, Vol. 1 No. 1, February 1985.

[6] Storch, R. L., Hammon, C. P., Bunch, H. M., and Moore, R. C., op. cit., pp. 163.

[7] Chase, R. and Aquilano, N., op. cit., pp. 93 1-936.

[8] Storch et. al., Process Analysis Via Accuracy Control, Revised, USDOT Maritime
~dministrationandTodd Pacific Shipyards Corporation, 1985.

[9] Corey, R., Industrial Marketing - Cases and Concepts, Fourth Edition, Prentice Hall Inc.,
Englewood Cliffs, NJ, 1991, pp. 7.

[lo] Chase, R. and Aquilano, N., op. cit., pp. 607-633.

[ l l ] Narasirnhan, S., McLeavey, D., Billington, P., op. cit., pp. 255-284,302-335,399-415.

[12] Ibid., pp. 282,485-496.

[13] Lee, T.E., Song, J.S.9 Im, J.C., Park, J.C., Jeong, D.S., and Lee, K.R., "Search-Based
Heuristic Algorithms for Basic Planning in a Large Shipyard," Journal of Ship Production,
Vol. 12, NO.4, NOV.1996, pp. 21 1-219.

[14] Pinedo, M., Scheduling Theory, Algorithms, and Systems, Prentice-Hall, Englewo~od
Cliffs, NJ, 1995.

[15] Bhattacharya, S., Production Planning and Control: An Integrated Approach, V i k i ~

Publishing House PVT Ltd. New Delhi, 1986, pp. 106-111.
[16] Mellichamp, J.M., and Love, R.M., "Production Switching Heuristics for the Aggregate
Planning Problem," Management Science, Vol. 24, No. 12, August 1978.

71 Minton, S., and Philips, A.B., "Applying a Heuristic Repair Method to the Hubble Space
Telescope Scheduling Problem," Innovative Approaches to Planning, Scheduling and
Control. Proceedings., Nov. 1990, pp. 2 15-219.

[18] Baker, E., Bodin, L., Finnegan, W., and Ponder, R., "Efficient Heuristic Solution to an
Airline Crew Scheduling Problem, AIIE Transactions, Vol. 11, June 1979.

[19] Chase, R. and Aquilano, N., op. cit., pp. 395-441,629-632.

[20] Narasimhan, S., McLeavey, D., Billington, P., op. cit., pp. 272-279,329-331.

[21] An Integrated Approach in Production Planning and Scheduling, Springer-Verlag, 1983,

pp. 1.1-1.4.3.

[22] Keith, E.G.,

"Operator Scheduling," AIIE Transactions, Vol. 11, No. 1, March 1979.

[23] Chase, R. and Aquilano, N., op. cit., pp. 395-441,541-569.

[24] Narasimhan, S., McLeavey, D., Billington, P., op, cit., pp. 585-613.

[25] Alkaner, S., and &per, LA., "Ship Production Scenario Analysis Using Multiple
Operations Research Techniques," Transactions, 8th International Conference on
Computer Applications in Shipbuilding, Sept. 1994, pp. 8.85-8.97.

[26] Amemiya, T., "Production Planning System for Opparna Shipyard," 8th International
Conference on Computer Applications in Shipbuilding, Sept. 1994, pp. 12.3-12.14.

[27] Kuhlmann, T., Marciniak, Z., and MaOow, C., "Integrated Coordination Modules for the
Shipbuilding Industry," Transactions, 8th International Conference on Computer
Applications in Shipbuilding, Sept. 1994, pp. 12.39-12.53.

[28] Tzannatos, E., and Markakis, K., "Computer-Aided Planning of Shipbuilding Operations,".
Transactions, 8th International Conference on Computer Applications in Shipbuilding,
Sept. 1994, pp. 12.67-12.77.

[29] Filipic, B., and Srdoc, A., "Task Scheduling and Resource Management in Ship Repair
Using a Genetic Algorithm," Transactions, 8th International Confirence on Computer
Applications in Shipbuilding, Sept. 1994, pp. 15.17-15.28.

1301 Gribskov, J., "A Group Technology Approach to Master Scheduling of Shipbuilding
Projects," Journal of Ship Production, Vol. 5, No. 4, Nov. 1989, pp. 249-255.
[3 11 Neumann, R.J., and McQuaide, D.J., "Application of PC-Based Project Management in an
Integrated Planning Process," Journal of Ship Production, Vol. 8, No. 4, Nov. 199:2, pp.
191-204.

[32] Neumann, R.J., "Network Scheduling Development in an MRP I1 Environment," Jburnal

of Ship Production, Vol. 10, No. 4, Nov. 1994, pp. 223-232.

[33] Correll, J.G., and Edson, N.W., Gaining Control: Capacity Management and Schelduling,
Oliver Wright Ltd. Publications Inc., Essex Junction, VT, 1990, pp. 60-63.

[34] Chase, R. and Aquilano, N., op. cit., pp. 694-750.

1351 Narasirnhan, S., McLeavey, D., Billington, P., op. cit., pp. 350-394.

[36] Neumann, R.J., op. cit.

[37] Scott., B., Manufacturing Planning Systems, McGraw-Hill, New York, 1990, pp. 1.27-148.

[38] Chase, R. and Aquilano, N., op. cit., pp. 906-930.

[39] Narasimhan, S., McLeavey, D., Billington, P., op, cit., pp. 561-584.

[40] Goldratt, E., The Haystack Syndrome - Sifting Information Out of the Data Ocean, North
River Press, Inc., Croton-on-Hudson, NY, 1990, pp. 163-262.

[41] Rack, F.H., "Significantly Reduced Shipbuilding Costs Through Constraint(s)

Management," The National Shipbuilding Research Program 1989 Ship Production
Symposium. Proceedings., Sept. 1989, pp. 17-1 - 17-17.

[42] Chase, R. and Aquilano, N., op. cit., pp. 787-817.

[43] Narasimhan, S., McLeavey, D., Billington, P., op. cit., pp. 508-512.

[44] Suresh, J., "A Simulation-Based Scheduling and Management Information System for a
Machine Shop," Interfaces, Vol. 6, No. 1, November 1975.

[45] Arnemiya, T., 'Process Planning System of CIM for Shipbuilding," Computer Applications
in the Automation of Shipyard Operation and Ship Design, VII, 1992, pp. 139-150.

[46] Bozenhardt, H.F., "Managing Complexity: A1 in Batch Plant Scheduling," - -

Handbook of
Expert Systems in Manufacturing, McGraw-Hill, New York, 1991, pp. 152-161.

[47] Chorafas, D., Expert Systems In Manufacturing, Van Nostrand Reinhold, New Yalrk, 1992,
pp. 199-212.
[4%] Maus, R. and Keyes, J., Handbook of Expert Systems in Manufacturing, McGraw-Hill,
New York, 1991, pp. 152-16 1,
[49] Butler, J., and Ohtsubo, H., "A Distributed Problem Solving Based Model for Shipyard
Scheduling," Transactions, Computer Applications in the Automation of Shipyard
Operation and Ship Design, VII, 1992, pp. 3 13-322.

[50] Parunak, V., Baker, A., and Clark, S., "The AARIA Agent Architecture: An Example of
Requirements-Driven Agent-Based System Design," http:llwww.iti.orgl, July 1996.

[51] Shine, J., "Herd Mentality," Wired, Vol. 4, No. 6, June 1996, pp. 98-104.

[52] Alfeld, L., The Virtual Shipyard - Simulating the Ship Construction Process, NAVSEA and
Advanced Marine Enterprises, September 1992.

[53] Spicknall, M., UMTRI Marine Systems Division Final Report: Evolution of Shipbuilding
Simulation Tools Using Realistic Shipyard and Product Contraints, NSWC-CD and
Westinghouse Machinery Technology Division, January 1995.

Some Other Relevant Information Resources

Shipbuilding-Specific Resources

Chirillo, L.D., Flexible Production Scheduling System, National Shipbuilding Research Program,
NSRP 0238, April 1986.

DeVries, R.L., "Computer-Aided Process Planning: A Path to Just-in-Time Manufacturing for

Shipyards," Journal of Ship Production, Vol. 4, No. 3, Aug. 1988, pp. 197-215.

DeVries, R.L., Allen, M.A., Jasper, M.L., and Barley, F.J., Computer Aided Process Planning
for Shipyards, National Shipbuilding Research Program, NSRP 0266, August 1986.

Koga, K., Ito, K., Nakai, Y., and Fujita, J., "Intelligent Design System of CIM for Shipbuilding,"
Transactions, Computer Applications in the Automation of Shipyard Operation and Ship
Design, VII, 1992, pp. 323-333.

Lazarus, P., "Reorganizing the Manufacturing Plant," Professional Boatbuilder, No. 29,
JuneIJuly 1994, pp. 22-33.

Minemura, T., "Scheduling Model of CIM for Shipbuilding," Transactions, 8th International
Conference on Computer Applications in Shipbuilding, Sept. 1994, pp. 12.25-12.37.

Nakarnura, M., Horiuchi, K., and Minemura, T., "Scheduling System of CIM for Shipbuilding
Applied Product Model, Process-Equipment Model and Operation-Resource Model,"
Transactions, Computer Applications in the Automation of Shipyard Operation and Ship
Design, VII, 1992, pp. 151-163.
Park, K.; Lee, K.; Park, 8.; Kim, S., "Modeling and Solving the Spatial Block Schedu1in;g
Problem in a Shipbuilding Company," Computers and Industrial Engineering, Val. 30, No.
3, July 11996, pp. 357-364.

Powell, P.C., and Zigelman, C.I., "Formal Manufacturing Approaches to Modem Shipbuilding,"
Journal of Ship Production, Vol. 5, No. 3, Aug. 1989, pp. 188-199.

Spicknall, M.H., "Developing and Using an Expert System for Planning the Production of
Structural Piece-Parts," Journal of Ship Production, Vol. 8, No. 3, Aug. 1992, pp. 1613-183.

Storch, R.L., and Chirillo, L.D., "The Effective Use of CAD in Shipyards," The National
Shipbuilding Research Program 1992 Ship Production Symposium. Proceedings., Sept.
1992, pp. 2B2-1 2B:2-10.
*-

Storch, R.L., and Hills, W. "Computer-Aided Manufacturing in Small Shipyards: A U.S. and
U.K. Comparative Study," Journal of Ship Production, Vol. 11, No. 2, May 1995, PI>.8 1-89.

Non-Shipbuilding Resources

White Papers From the American Production and Inventory Control Society, APICS
@ttp://lionhrtpub.com/apics/)

Baker, T. and Cleaves, G., "World-Class Performance Through Improved Planning anld
Scheduling Integration," October 1991.

Gilman, A, "Is MRP For You?," March 1995.

Gilman, A., "Interest in Finite Scheduling is Growing.. .Why?," August 1994.

Hunter, M., "How To Make A Realistic Production Plan," June 1992.

Kiran, A. and Willingham, T., "Simulation: Help For Yow Scheduling Problems," August
1992.

Marche, S., "The Seven Key Elements to Selecting a Successful Production Schedu1in.g
System," March 1996.

McKaskill, T., "There is a Place for Expert Systems in Finite Scheduling," August 1994.

Merithew, C., "How Do You Plan In a Mixed Mode Environment?," October 1993.

Nichols, J. and Thompson, M., "Planning for Real World I'roduction," August 1994.

Pritsker, A. and Snyder, K., "Simulation for Planning and Scheduling," August 1994.

Proud, J., "More Than Just Numbers! Master Scheduling Management Issues," March 1995.
 Schutt, J., "The Quiet Revolution: Integrated Systems For Operations Planning," November
1993.

Sharrna, K. and Wilson, J., "What's Wrong With Finite Capacity Scheduling?," March 1995.

Suri, A., "Master Production Schedule: The Driver of Planning and Control Systems," July
1992.

Taylor, S. and Bolander, S., "The Rise of Finite Capacity Scheduling," July 1996.

Thompson, M., "Computer Simulation Drives Innovation In Scheduling," February 1993.

Thompson, M., "Why Finite Capacity? Things To Consider When Selecting Scheduling
Software, June 1992.

Voet, M. and Dewilde, P., "Choosing a Scheduling package: Thirty Systems Evaluated in
Worldwide Survey," November 1994.

Wyman, F., "Thinking About Simulation-Based Scheduling? Here's What You Ought To
Know," November 1993.

Zweben, M., "Intelligent Agents: Planning and Scheduling in Real Time," March 1995.

White Papers From the Institute for Operations Research and the Management Sciences,
INFORMS @ttp://www.infoms.org/)

Adenso-Diaz, B.and Laguna, M., "Modeling the Load Leveling Problem in Master Production
Scheduling," 1996.

Arosio, M. and Sianesi, A,, "A Heuristic Algorithm for Master Production Schedule
Generation With Finite Capacity," 1993.

Chu, S., "A Mathematical Programming Approach Toward Optimized Master Production
Scheduling," 1995.

Chung, C. and Krajewski, L., "Planning Horizons For Master Production Scheduling," 1984.

Das, S. and Sarin, S., "An Integrated Approach to Solving the Master Aggregate Scheduling
Problem," 1994.

Das, S., "Master Scheduling With Incremental Capacity Allocation and Rolling Horizon,"
1993.

Giesberts, P., "Master Production Scheduling: A Function-Based Approach," 1991.

Hong, S. and Maleyeff, J., "Production Planning and Master Scheduling With Spreadsheets,"
1987.
 King, B. and Benton, W., "Alternative Master Production Scheduling Techniques in an.
Assembly-To-Order Environment," 1987.

Sridhara, V. and Berry, W., "Freezing the Master Production Schedule Under Rolling Planning
Horizons," 1987.

Ware, N. and Fogarty, D.,"Master Schedule/Master Production Schedule: The Same Or

Different?," 1990.

Zhao, X. and Lee, T., "Freezing the Master Production Schedule for Material Requirements
Planning Systems," 1993.

Other Documents

Bensoussan, A., Crouhy, M., and Proth, J.M., "Mathematical Theory of Production Planning,"
Advanced Series in Management, Vol. 3,1983, pp. 2-15.

Eilon, S., "Five Approaches to Aggregated Production Planning," AIIE Transactions, Vol., 7,
No. 2., June 1975.

Fraser, J., "Finite Scheduling and Manufacturing Synchronization: Tools for Real Plant
Productivity" (with software buyer's guide), IIE Solutions, Sept. 1995.

Glover, F., Jones, J., Karney, D., Klingman, D., and Mote J., "An Integrated Production,
Distribution, and Inventory Planning System," INTERFACES, Vol. 9, No, 5, Nov. 1979.

Jain, S.K., "A Simulation-Based Scheduling and Management Information System for a
Machine Shop," INTERFACES, Vol. 6, No. 1, Pt. 2, Nov. 1975.

Kahaner, D.K., "Computer Applications, Production Engineering China,"

http:llwww.atip.or.jp/, June 1995.

Peng, W., Knowledge-Based Master Production Scheduler for the Computer-lntegratt!d

Manufacturing System, Ph.D. thesis, http:/lwww.uta.edu/ie/peng.html, 1989.

Proud, J.F., Master Scheduling, a Practical Guide to Competitive Manufacturing, Olivier Wight
Publications Ltd., John Wiley and Sons, Essex Junction, VT.

Smith, J., "TOC and MRPII, From Theory To Results,"

http://www.rogo.com~caclJJSmith.html, May 1994.

Internet Resources
Organizations are identified in parentheses,

Informational URLs-
http:llainet.comlopticsus/jit.htm (Optimal Planning Techniques, Inc.)
 http://lionhrtpub.com/apics/BG/bgo.html (American Planning and Inventory Control
Society)
r http://lionhrtpub.com/apics/BG/BGPLMES.html(American Planning and Inventory
Control Society)
http://tamcam.tamu.edu/Papers/scheduling.htm(Texas A&M Industrial Engineering
Department)
r http://tamcm.tamu.edu/SBSC/index.htm#Target Environment (Texas A&M, Industrial
Engineering Department)
http://www.cen.uiuc.edu/-ie26 l/ie262/notes/MRP/IE373-MRP.htm1 (University of Illinois,
College of Engineering)
http://www.cs.cmu.edu/-SWWManagemnt/html/mod2/mod23.html (Carnegie Mellon
University, School of Computer Science)
r http://www.ececs.uc.edu/-mnoschan/agents.html (University of Cincinnati, Department of
Electrical and Computer Engineering, and Computer Science)
r http://www.gmcs.co.uk/ (GMCS, Great Britain)
http://www.industry.net/c/orgindex/apics(1ndustry.Net)
e http://www.pmi.orgl (Project Management Institute)
r http://www.primavera.com/sources/index.hl (Primavera, Inc.)
http://www.rogo.com/cac/JJSmith.html("Crazy About Constraints" - Information on the
Theory of Constraints)
http://www.uta.edu/ie/index.htm (University of Texas, Arlington, Industrial and
Manufacturing Systems Engineering Department)

System and Software Vendor URLs-

http://merlion.singnet.com.sg:80/-gilcresmh (Micro-MRP, Inc.)
r http://ra.pyrmid.com/solutionsPartner~Directo~~tm~O755.h~1 (Siemens)
http://www.agamasoft.com/ (Agarna Software, Inc.)
http://www.aimcom.com/ (AIM Computer Solutions, Inc.)
http://www.ainet.com/opt/ (Optimal Planning Techniques, Inc.)
http://www.alliancemfg.com/ (Alliance Manufacturing Software)
http://www.amsoftware.com/ (American Software)
r http://www.axiscomp.com/ (AXIS Computer Systems, Inc.)
http://www.baan.com/33SolutionslApplications/l Mfg/default.htm (The Baan Company)
http://www.batchrnaster.com:80imrp.html ( ~ a t c G a s t e Software
r Corp.)
r http://www.cincom.com~products/manufac.html (CINCOM)
r http://www.dataworks.com/ (Dataworks Corp.)
r http://www.dcdcorp.com/ (DCD Corp.)
http://www.dg.com~Solutions-Directory/prdOO665.h1 (Data General Corp.)
r http://www.distinction.com/ Distinction Software, Inc.)
http://www.exel.co.uk~pagesMain~productEFACSNThtm (Exel Computer Systems)
http://www.fs.com:80/products/specsheets.h (Fourth Shift Corp.)
r http://www.i2.coml (i2 Technologies)
e' http://www.inmass.com/ (INMASSRVIRP Software)
http://www.jdedwards.com/ (J. D. Edwards, Inc.)
http://www.manugistics.cod (Manugistics, Inc.)
http://www.mapics.corn:80/app/3mpsp.html (MAPICS, Inc.)
http://www.maxam.com/ (Maxam Holdings, Inc.)
hrcp://www.mimerva-is.com (Minerva)
http://www.monitor-toronto.com/ (Monitor Systems of Toronto)
http://www.mrp9000.com/ (Intuitive Manufacturing Systems, Inc.)
http://www.natlbus.com/pcms.htm (National Business Consultants)
http://web.onramp.ca/ (Onramp Network Services, Inc.)
http://www.osas.com:80/products/addon~specslhtml (Open Systems Holdings, Inc.)
http://www.paragonms.cod (Paragon Management Systems)
http://www.peoplesoft.coml (Peoplesoft, Inc.)
http://www.plnlrocktec.coml (Plymouth Rock Technologies, Inc.)
http://www.primavt:ra.com/ (Primavera, Inc.)
http://www.ramco.coml (RAMCO Systems Corp.)
http://www.rtexpertsource.com/dbmB.cgi?es+Specialition+Production Planning (RT
Expert Source)
http ://www.sap.cod (SAP)
http://www.shivasoft.coml (Shivasoft, Inc.)
http://www.sm.com/ (Systems Modeling Corp.)
http://www.ssax.comlbpcs/mmm-mrp.htm (System Software Associates, Inc.)
http://www.synques;t.coml (Synquest, Inc.)
http://www.sysprousa.cod (Syspro Impact Software, Inc.)
http://www.telesismfg.com/ (Telesis MFG)
http://www.thru-put.com/ (ThruPut Technologies)
http://www.tiac.net/'users/sailboat/(MCS-3 Software)
http://www.tipsol.cod (Total Integrated Product Solutions)
http://www.tiwcorp.coml (TIW Technology Corp.)
http://www.txbme.com~manufacturingl (TXbase)
http://www.tyecin.coml (TYECIN Systems, Inc.)
http://www.uk.mdis.codproducts/chess/chessmasterscheduling.hl (MDIS)
http://www.usersol.com/us1002.html (User Solutions, Inc.)
 Transportation
Research Institute