SHIP PRODUCTION COMMITTEEFACILITIES AND ENVIRONMENTAL EFFECTSSURFACE PREPARATION AND COATINGSDESIGN/PRODUCTION INTEGRATIONHUMAN RESOURCE INNOVATIONMARINE INDUSTRY STANDARDSWELDINGINDUSTRIAL ENGINEERINGEDUCATION AND TRAINING

THE NATIONALSHIPBUILDINGRESEARCHPROGRAM

August 1990NSRP 0320

1990 Ship Production Symposium

Paper No. 1B-1:Shipyard Modelling -- Approach toObtain Comprehensive Understandingof Functions and Activities

U.S. DEPARTMENT OF THE NAVYCARDEROCK DIVISION,NAVAL SURFACE WARFARE CENTER

Report Documentation Page Form ApprovedOMB No. 0704-0188

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering andmaintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, ArlingtonVA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if itdoes not display a currently valid OMB control number.

1. REPORT DATE AUG 1990

2. REPORT TYPE N/A

3. DATES COVERED -

4. TITLE AND SUBTITLE The National Shipbuilding Research Program, 1990 Ship ProductionSymposium, Paper No. 1B-1: Shipyard Modelling -- Approach to ObtainComprehensive Understanding of Functions and Activities

5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) Naval Surface Warfare Center CD Code 2230-Design Integration ToolsBldg 192, Room 128 9500 MacArthur Blvd, Bethesda, MD 20817-5700

8. PERFORMING ORGANIZATIONREPORT NUMBER

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

11. SPONSOR/MONITOR’S REPORT NUMBER(S)

12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release, distribution unlimited

13. SUPPLEMENTARY NOTES

14. ABSTRACT

15. SUBJECT TERMS

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT

SAR

18. NUMBEROF PAGES

21

19a. NAME OFRESPONSIBLE PERSON

a. REPORT unclassified

b. ABSTRACT unclassified

c. THIS PAGE unclassified

Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18

DISCLAIMER

These reports were prepared as an account of government-sponsored work. Neither theUnited States, nor the United States Navy, nor any person acting on behalf of the UnitedStates Navy (A) makes any warranty or representation, expressed or implied, with respectto the accuracy, completeness or usefulness of the information contained in this report/manual, or that the use of any information, apparatus, method, or process disclosed in thisreport may not infringe privately owned rights; or (B) assumes any liabilities with respect tothe use of or for damages resulting from the use of any information, apparatus, method, orprocess disclosed in the report. As used in the above, “Persons acting on behalf of theUnited States Navy” includes any employee, contractor, or subcontractor to the contractorof the United States Navy to the extent that such employee, contractor, or subcontractor tothe contractor prepares, handles, or distributes, or provides access to any informationpursuant to his employment or contract or subcontract to the contractor with the UnitedStates Navy. ANY POSSIBLE IMPLIED WARRANTIES OF MERCHANTABILITY AND/ORFITNESS FOR PURPOSE ARE SPECIFICALLY DISCLAIMED.

THE SOCIETY OF NAVAL ARCHITECTS AND MARINE ENGINEERS601 Pavonia Avenue, Jersey City, NJ 07306

Paper presented at the NSRP 1990 Ship Production Symposium,Pfister Hotel, Milwaukee, Wisconsin, August 21-24.1990

Shipyard Modelling-Approach to ObtainComprehensive Understanding of Functionsand Activities 1B-1

Joachim Brodda, BREMER VULKAN AG and Bremen Institute of Industrial Technology (BIBA)

NOMENCLATURE

CAM

CASECIMCNCDNC

ESPRIT

GRAI

ICAM

IDEF

NIAMNIDDESC

PDESR + DSTEP

CAD Computer Aided DesignComputer Aided Manufacturing

CAP Computer Aided PlanningComputer Aided Software EngineeringComputer Integrated ManufacturingComputer Numerical ControlDirect Numerical Control

ERA Entity Relationship ApproachEuropean Strategic Programme for Research and Development InformationTechnologyGroupe de Recherche en AutomatisationIntegrielIntegrated Computer AidedManufacturingICAM Definition

IT Information TechnologyNijssen Information Analysis MethodNavy/Industry Digital Data ExchangeSpecification CommitteeProduct Data Exchange SpecificationResearch and DevelopmentStandard for the Exchange of ProductModel Data

ABSTRACT

The application of Computer Manufacure (CIM) to one-of-a-kind

Integratedheavy engineering

industries IS one of today’s most challenging tasks forthe industry and Information Technology (IT)vendors. Because of the complextity of the shipbuildingprocess, as a

?typical representative of this kind of

Industry, available solutions are not yet satisfactory.Even the basic theory of current CIM sytems does notfit many aspects of requirements for the shipbuildingindustry. The application of CIM tools, which arebased on ideas of line production and have a strictsequence of all tasks, have very often failed in the past,mainly because of a lack of flexibility. Therefore, basicwork needs to be performed, starting with a detailedanalysis of today’s situation and leading to a carefuldevelopment for the realThis paper deals with the problems

requirements of tomorrow.lems resulting from the

exceptional nature of shipbuilding for the applicationof CIM elements. The approach chosen for modellingof the production process, illustrated by some examplesfrom actual Reseaarch and Developmentprojects, is described. An evaluation of

(R+D)the benefit of

structured shipyard complementary

modelling and a look atR+ D actions concludes the paper.

INTRODUCTION

Heavy industries, concerned with "one-of-a-kind". production, are typically described throughcombined workshop manufacturing and remoteconstruction site assembly, and have not yet fullybenefited in the application and use of modeminformation technologres.

The design and manufacturing of large multisystems intergrating products (ships, offishore structures,plants, factories) is a complex and long term activitycovering a whole range of possible engineering andworking activities. The manufacturing,subdivided mainly into three parts: fabrication using

process can be

standard and non-standard raw materials; theprefabrication of modules, including pre -assembly withother prefabricated orfinal outfitting activities. In

purchased products; and the this complex manufacturing

process, all production activities can normally be foundat one production site.

The manufacturing process is construction siteoriented with its signification cant amounts of specializedworkshop production. Thedesign, planning and

typical overlapping ofmanufacturing means that,

compared to mass production with line character, therehas been a delay in the development of tools for thesemore complex requirements.

To come to a CIM solution in this kind ofindustry, a lot of additional basic work is requiredwhich can nevertheless be set up based on commonstandardization of definitions and rules.

Over thehas not been l

past ten years the shipbuilding industryslow in adapting the latest Computer

Aided Design (CAD) techniques. Similarly, althoughthe use of numerically controlled (NC) equipment,, e.g.for flame cutting, is commonplace. In this context thereis already some computerized intergration of CAD andComputer Aided Manufacturing CAM). Some linksrare also existing between CAD and Computer AidedPlanning (CAP) and Production Planning and Control(PPC) Systems through the generation and completionof bills of materials.

However, the majority of systems in use aremostly ‘island’ solutions which are supporting work inspecial applicationSystems the

fields. Below the level of PPC- information flow is mainly paper based, or

on a person-to-person level.

1B-1-1

Before starting into the development o f i d e a sfor CIM applications, the need for aof the

detailed analysisresent situation of one-of-a-kind production in

genera and shipbuilding in particular is evident. Theanalysis should cover different viewpoints of theprocess, such as manufacturing functions, processplanning and control functions, organizational aspects,resources and information links. The use of formalisedmodelling methods even for analysis purposes providesa good basis for the development of requirement the future system, i.e. an architectural reference

form o d e l .

This model defines the complete set of functionalities,single activities and interrelations required for theproduction process. CIM elements and in particularalso the remaining manual tasks as intergratedelements can be identified or defined on the basis ofthis reference model.

Organizational changes, the development ofCIM elements, and the specification of interfaces asmain element can be prepared. Any changes withrespect to the today’s situation can be documented.

Models based on common descriptive languagestherefore provide a good opportunity for the discussionbetween system users and system developers. Themore complex the task, and the less clear activitysequences are, the more important the use offormalised modelling techniques is. The originalshipbuilding process combined with its manifoldexternal dependencies represents a challengingproduction process to be described.

THE CHARACTER OF SHIPBUILDING

The shipbuilding process has been chosen for anumber of principal actions within European R + D.As reference products of one-of-a-kind manufacturingships provide goodinvestigations.

opportunities for basicCompared with mass production,

significant differences can be identified as:

-- tremendous influence on design andmanufacturing by the customers;

-- complex , complicated and multi-stageproduction process withinterdependencies;

high

-- combined manufacturing principles atone site;

-- use of universal equipment;

-- craft skills are of vital importance for theassembly process;

-- long term order throughput;

-- character of products under contractchanges;

-- significant overlapping ofplanning and manufacturing;

design,

-- hostile working environments;

-- final product definition only aftercontract signing possible;

1B-l-2

-- decisions with high relevance must betaken on basis of uncertain, stochasticinformation;

-- product value is very high;

-- order throughput times are very long;

-- product size in volume and weight is veryhigh.

In spite of being incomplete, this list gives somemajor reasons for the difficulties in the utilization ofadvanced information technologies for shipbuildingwhich are originally designed for mass production.

Some of the points will be illustrated in the following.

- Steel Hull Structure- Accomodations

- Systems to Distribute- Sea and Fresh Water

- Systems for- Energy Generation

:Steam and and Condensate

- Fuel and Diesel Oil

‘Figure 1: Ships-Multi Systems Products [10]

Ships, as unique and ambitious technicalobjects, contain numerous and different technicalsystems (Figure 1). These different systems require arelated number of various skills and manufacturing

principles. Different types of ships require differentloads in typical work trades which can lead toconsiderably divergent loads. Figure 2 shows anexample of the loads for four typical ship types.Depending on the type of shipyard - with high4 or lowlevel of self-fabrication of parts and components - thisalso effects the collaboration with subcontracters orsuppliers. Even in case of specialization of a few shiptypes the flexibility of work trades and equipment mustbe kept. Therefore all facilities of the shipyard have tobe designed for the widest range of products allowingnearly the same level of productivity and quality for thedifferent production cases. In this context it isunderstandable that for shipbuilding all activitiesworking towards an optimisation and integration ofdesigning, planning, manufacturing, and assembly areof vital interest. Especially, methods and tools forplanning, monitoring and control of the uncertainprocess must be a matter of special consideration.

Duringthe design office

the preparation of offers for customers must be able to provide information

of high accuracy for the calculation of requiredmaterial, needed resources, different loads of

• Steel Hull q Steel Outfitting, Accommodation

q Engine Installation, Pipes Electrical Plant

ž Auxillary Plant, etc.

Figure 2: Different Ship Types WorkloadDistribution or Main Work Trades [10]for

t Decision Relevance -

IOrder ISart AssStart Fabrircation

emblyDelivery

Figure 3: Decision Relevance Compared with Data

worktrades and time schedules. In this early stage of apotential order, decisions have to be made on the basisof forecasted figures which are of importance for thewhole order throughput time. The preliminary design,without any detailed definition of the product, providesthe more or less uncertain basis for those forecasts anddecisions. The success or failure of an order for acompany therefore depends highly on the skill andknowledge of the decision-making people whocompensate for the lack of information in thisstage of a customers request with their experience.

early

Not only in this initial phase must missing databe compensated for by experience, but even in the production phase, mainly in the assembly process,workers make decisions about construction and designsolutions, e.g. the routing of pipes. Therefore, ship, orcompareable objects in size and complexity, do notnecessarily need a 100% description by data. As longas a certain level of experience can be held atshipyards, the maximum level of data can be limited.Nevertheless, today's situation does not seemsatisfactory.

dTherefore, one of the tasks to be

performe is to find out the suitable level of data to beproduced by advanced information technologies andthe maximum gap to be levelled through humanexperience in the different stages of the orderthroughput. Figure 3 illustrates this coherence.

A reason for the need to balance the lack ofdata through experience can be seen in the significantoverlapping of the different order throughput phases.Figure 4 shows this overlapping compared with thesituation in mass or series production. Because some

I

Figure 4: Different Overlapping of Main Activitiesfor Serial Production and One-of-a-RindProduction

1B-1-3

activites have to start before the required data hasbeen produced, the ‘realtime’ generation of workinginstructions such as ‘what to do’ and ‘how to do it’ isrequired. Those working instructions very often areproduced by the shopconsulting the design office

floor responsible withoutand higher level planning

instances. The trust in the self responsibility and.experience of these people by the company must behigh. In spite of the danger this system naturallycontains, advanced organizational principles, highlysupported by modern information technologies, shouldcarefully consider this flexibility of today’s system as anasset of today, as well for tomorrow. Therefore a future(CIM) system must consider even the human role as aneminent element of the process.

The initial phase of the contract, i.e. thepreliminary design, bases all calculations on thefunctional structure of the ship. The product will be

designed from the system point of view. Constructionalgroupings (e.g. ballast water system) normally areapproved structures of the product traditionally grownor tailor made for particular shipyards. This functionalview of the system “ship” provides a good structure fordifferent tasks in design and pre-calculations(functional structure). Because of the separateconsideration of all constructional groups, resources,loads, andcan be

possible due dates, for this potential ordercalculated on this basis and preliminarily

mapped into plans, cost centers and financial(cost/ quantity structure). Based on

planning

capabilities and and constraints (e.g. lifting capacities,,.manufacturing

space) the ship will be divided into manufacturing specific uspecificmist ructure) which provides

the backbone for the bills of material for structure andfurther scheduling purposes (Figure 5).

Figure 5: Threefold view on the Product Ship[BIBA]

Modern ship production techniques, likebuilding big blocks with high quantities of pre-outfitting before assembling the ship within drydocks,defintely need a suitable manufacturing structure. With

Figure 6: Integration of Manufacturing Structureand Functional Structure [BIBA]

1B-1-4

the ongoing subdivision of the ship body the functionalgroups must be subdivided similarly and distributed torelated blocks and sections (Figure 6). This means therelated bills of material, job cards and other planningdocuments must be merged in a suitable way whichcreates difficulties for conventional planning methodsand tools.

The intention to extend the level of preoutfittingand to reduce the time spent at the final building placeby the object makes it more important to think aboutthe right structures and their interlinkages.

The different stages of the shipbuilding processcan be subdivided into 7 levels (Figure 7). Anadditional stage for section or unit conservation shouldbe considered between levels 3/4 or 4/.5. Because ofthe growing size and weight of the different objectsbetween levels 2/3 there is usually a point during steelassembly where a transition is necessary from “movingproduct to process” to “moving process to product".Therefore levels 0, 1 and 2 can be performed followingPworkshop production principles. On the other hand,levels 3 to 6 are mainly performed at construction sites.Figure 8 defines the different manufacturing principles.The three main parts of the production system - theworking object, the worker and the productionequipment - are defined as fixed or movable. In case ofconstruction site manufacturing, the working object isfixed, at least for a certain time. All equipment andworkers have to be moved to it. The main shipbuilding

processes can be easily identified as belonging to this kind of manufacturing principle. Different pre-

fabrication processes(e.g. pipes,

for steel and outfitting tradessheet metal, accomodation) can be

identified as workshop or even line productionoriented processes (Figure 9).

Final outfitting I

6Steel Shipbuilding Finaoutfitting O U T

- - - - - - - - - - - - - -

Pre - outfitting(4)

4 Unit (Blocks) Pre- outfittingErection (3)

- - - - - - - - - - - - - - - - _ _ _

Volume SectionErection

Pre - outfitting(2) I. ,

_ - _ - - - - - - - - - - - - - - - -

2 Flat SectionErection

Pre - outfitting(1)

- - - - - - - - - - - - - - - - - - _

GroupsErection I Pre - assembly

I- - - - - - - - - - - - - - - - - - -

I- t I-

Figure 7: Levels of the Shipbuilding Process

0 w P

0 0

WORKSHOP PRODUCTION

CONSTRUCTION SITEMANUFACTURING

0 = Working ObjektW = Worker 0 fixed

P = Production Equipment 0 movable

Figure 8: Definition of Manufacturing Principles

0

6

Figure 9: Assignment of Manufacturing Areas toManufacturing Principles

These workshop or line production orientedtasks accompany the erection/construction siteoriented processes throughout the whole constructiontime. Advanced shipyard concepts try to group thoseworkshops closely around the related outfittinglocations. For instance, Bremer Vulkan AGimplemented the so called Workshop Oriented ShipProduction Technology (WOST) concept which led tosignificant short cuts and reductions in outfitting costs.The idea was to minimize information links fromworkshop to the construction site by bringing outfittingintensive blocks under roof, close to the workshops.

1B-1-5

This concept was realized for the engine room sectionand the superstructure. The highly pre-outfitted blocksthen can be transferred to the dry dock by a gantrycrane. After integrating the blocks into the hull as afinal step of ship erection, the ship can be launchedimmediately.

Nevertheless some major tasks remain onremote construction sites. Because of hostileevironments, uncertain planning basis, unforeseeablechanges through late delivery of material or weatherinfluences, the monitoring and control of theproduction process is relatively difficult. Major reasonsfor these difficulties can be seen in communicationproblems of central planning instances with remoteworking people.

However the described character of theshipbuilding process shows some differences to otherindustry’s. At least it provides some reasons for theneed of some exceptional requirements and the needfor adapted CIM theories and tools which are not yetsatisfactorily provided by scientists and vendors.

MEANING OF CIM FOR SHIPYARDS

The integration of computers into the physical 2.manufacturing process and all related managementfunctions. for heay engineering industries in generaland for shipbuilding in particular , is a very challengingtask. First the question must be answered whatcomputer integration means for this type of industry. Itshould be considered that 80% of the working hours oreven more today are manual assembly tasks whichcan’t be controlled by NC devices. This number willnot significantly change within the near future.Therefore the industry highly de ends on the skill, theself-responsibilty and the flexibility of its workingpersonnel on the shop floor. Many naturally existingdeficits in craft manufacturing can be solved throughimprovements in organization, facilitating, and bettersupport for the handling of material and layoutimprovements.

Naturally all possibilities for manufacturingautomation should be considered. Solutions for thefabrication, e.g. of steel parts, are commonplace. Forlow level assembly tasks, mainly in the field of weldingof subassemblies, some promising tools are available.Nevertheless investments in these tools comprise somedifficult calculable risks and require corporatedecisions. Beyond the question of how it is useful tomechanize or automate physical manufacturingprocesses, the generation and use of information fortechnical and management purposes must be carefullyinvestigated. A process which depends highly ondecisions made by experience needs improvements ininformation provision. Developments in this field aremajor objectrves of CIM approaches for shipbuilding.

CIM approaches should always be seen as anoverall company strategy, giving more answers for“Integration” than just for “Computers” and“Manufacturing”. In this context, the definition of clearcorporate strategic targets for CIM is a must for everysingle company. Functions and activities of themanufacturing process have to be considered as havingvery close and manifold links to all other necessarytasks. This is particularly important to solve theintegration aspect of CIM. However, it is essential forCIM to cover the whole range of company

manufacturing activities. Partly implemented ‘islandsolutions’ or small groups of integrated systems mighthave improved the productivity of single companydepartments. However, the benefits of llan overa 1integration are greater than those from the sum of‘island solutions’.

The one-of-a-kind nature of the product (ships),including the related special demands formanufacturing process and management, lead to someextended and exceptionel requirements for CIM.Because of individual differences between companiesand a relativiely small market for IT vendors,combined with complex function for suitable softwareelements, many of these requirements couldn’t besatisfied in the past. The man-machine interfaces arean es ecially imortant factor in the existing craftskilled 8 .future.

dominated industry and will remain so in the

Thinking about movements towards CIM oreven Computer Integrated Enterprise (CIE) thefollowing field seems to be important in the future.

1.

3.

4.

5.

6.

Consequent use of 3D-CAD-systems withcomplete substitutionengineroom models.

of physical

CAM systems with complex links to CADfor NC path generation and simulation

(

e.g. Computer Numerical ControlledCNC) or Direct Numerical ControlledDNC) weldingrobots for first to third

stage assembly).

Integrated Computer Aided Engineering(CAE) and CAD systems allowing stagewise operations planning (for calculationpurposes and manufacturing planning).

PPC combined withdecentralised worktrade oriented controlcenters to serve three different levels ofplanning accuracy.

Object oriented production progressdevices using user friendly data capturing

Communication technology for externallinks and internal fixed, temporary andmovable requirements.

Neutral data base datastructure, and database management concepts for productdata, factory data and process data.

New concepts for data presentation topeople in the yard (e.g. in the field ofprogress monitoring).

Knowledge Based Systems (KBS)especially for rescheduling purposes.

The hope for develo ments and improementin all these different IT fieds are natural y based onl open systems and neutral data storage concepts. Inparticular, shipbuilding and similar heavy engineeringindustries will definitely benefit from these approaches.It is doubtful whether individual, temporary or ‘closedsolutions can ever be gained from one-off productmanufacturing.

THE NEED FOR MODELLING

To ensure that all single applications will fit intothe whole CIM infrastructure in the shipyard, anoverall CIM strategy must be found. Understanding theenterprise is essential before the future architecturescan be developed. Existing system strutures have to bemodelled in a compact and comprehensive form.Because of the natural interest to use generalapproaches it is necessary to think strategically and inlong terms. Conventional, manual and intuitiveapproaches of today are not satisfactory in this context.Especially, the coordination of physical processes withinformation systems regarding hierarchicalcommunication often exceed human imaginationbecause of their complexity. Therefore, formalizedtechniques should be applied to analyse existingenterprize functions, information/data structures,organization and resources. The development o f n e wstructures leading to a reference model or future CIM applications should be based on the same techniquesutilizing existing elements of the original structure.This approach provides a common language for the

participating architects, users, experts, non-experts andiT developers from the beginning, and throughout the

whole develo ment period. The model also provides agood basis or testing, simulation and cost-benefit-analysis of intended changes compared with theexisting system.

The performance of methods and tools to beutilized should further allow mapping the

d eprocesses

globally and on detailed levels. At east the definition,specification and design of soft- and hardware shouldbe supported. A three level approach to come to animplementation specification (Figure 10) follows in

project 66principle the particular derivation process of ESPRIT, . 8 ESPRIT is the “European Strategic

Programme for Research and Development inInformation Technology.” The objective of thiss projectwas to design an Open System Architecture (OSA) forCIM and to define a set of concepts and rules tofacilitate the building of future CIM systems.

The (future) reference model should be definedthrough careful analysis with transformation of the as ISsituation combined transformation anddevelopment into an ideal ‘should be’ scenario. Thecomparison with possible and available organizationaland IT solutions lead, at least through intermediatestages, to an integrated implementation specification.The reference model updated through theimplementations taken provides a basis for continuedresearch and definition of an advanced CIM design.

MODELLING APPROACH

Several methods and tools for the differententerprise modelling tasks have been developed in thepast. Those tools are often based on Computer AidedSoftware Engineering (CASE) tools and are utilizedfor the different modelling tasks. Because of therequirements of shipbuilding, including lots ofdecisions based on experience, and because of the

Especial interest in the planning and control sector, thesPRIT Project No. 2439 ROCOCO (Real Time

Monitoring and control of construction SiteManufacturing) decided to follow the GRAIde Recherche en Automatisation Integriel)

(Grouhapproac

for modelling and methods. The ROCOCO project islead by Bremer Vulkan AG and involves 4 more major

Figure 10: Three Level Approach for FactoryModelling

European shipyards (Chantiers de L’Atlantique,Fincantieri, Eleusis Shipyard, Masa Yards). At least 11partners from Europe comprising research institutes,universities and IT vendors beneath the shipyards areforming the consortium.

To provide the basis for the tool developmenttasks, and as a major part of the project, a referencearchrtecture customized to the intended applicationarea must be developed.

Following the GRAI a preachreference model will be sub divided

the structuredivided into three sub-

sytems (Figure 11). The operational sub-systemdescribes the physical manufacturing (and design)functions and has the role of transforming, rawmaterials (and design orders) into end-products (andtechnical data).

The decision sub-system, also called theproduction and design management sub-system aims,to control the operatronal sub-system in order or reachthe economic targetsaccount (Figure 12)

while taking constraints into The information sub-system links

the two previous sub-systems and aims to supply andmemorize the information, restitute and process it.From the various applicable methods and approachesIDEF 0 (see next subchapteroperation sub-system and

) has been choosen for theGRAI for the decisional sub-

system. For the informational sub-system an EntityRelationship Approach (ERA) has to be applied. The

1B-1-7

decision for a particular tool for information modellinghas not been finally taken. Tests with several tools (e.g.IDEF 1 or the Nijssen Information Analysis Method(NIAM)) in context with other projects are under way.

Figure 11: Reference Model-GRAI Approach [6]

follow up Information

Itechnical data

manufacturing orders

- and Design design ordersManagement

l

Figure 12: Assignment of Methods to Sub-Systems

IDEF Methodology

Originally the methodology has its roots in theUS Air Force Program for Integrated Computer AidedManufacturing (ICAM). This program identified theneed for better communication and analysis betweenthe people involved in improving manufacturingproductivity. To satisfy that need the ICAM programdeveloped the IDEF (ICAM DEFinition) methods to

faddress particular characteristics of manu acturing.

The approach was to use the methods toproduce models which would provide a basis fordefining where changes to the manufacturing processwould result in improvements to manufacturingproductivity.

whichIDEF comprises three modelling methodologies

is use graphically characterize manufacturing. IDEF 0to produce a “functional model”, which is a

structured representation of the functions of a systemand of the information and objects which interrelatethose functions. IDEF 1 is used to produce an“information model” which represents the structure andsemantics of information within the system. IDEF 2 isused to produce a “dynamic model” which representsthe time varying behavioral characteristics of thesystem.

IDEF 0, which has been applied here,. consistsof techniusegraphical

for performing system analysis and alanguage for applying these techniques. The

graphical language is limited to a set of basiccomponents with which the analyst or designer cancompose structures of any size. These basiccomponents include boxes and arrows.

The IDEF 0 diagrams in a model are or organizedin a hierarchical and modular “top-down” fashionfashionshowing the breakdown of the system into itscomponent parts. Application of IDEF 0 starts with themost general or abstract description of the system to beproduced. If this description IS contained m a single‘module,” represented y a box, that box is brokenbdown into a number of more detailed boxes, each ofwhich represents a component part. The componentparts are then detailed, each on another diagram. Eachpart shown on a detail diagram is again broken down,and so forth, until the system is described to anydesired level of detail. Lower level diagrams (children),then, are detailed breakdowns of higher level diagrams(parents). The place of each diagram in a model isindicated by a “node-number”, derived from thenumbering of boxes. The boxes within the diagramsrepresent functions or activities in the hierarchy namedby verbs connected by arrows representing therelationship (objects, information) labeled with nouns.

Figure 13 maps the application of themethodology for the ROCOCO project. Two differentdiagrams of the same level showing the same activityboxes describe first the material flow and second theinformation flow. This approach has been chosen forthe ROCOCO project to keep the diagrams simpler.The lowest level diagrams are clear-cut single activitiesdescribing all inputs, outputs, controls and resources(material and information related). This gives acomplete figure for these single activities and can beconsidered as kind of generic modelling elementsforming the basis for the future definition of othermodels. A textual explanation diagram behind theoriginal diagram gives the opportunity for additionalcomments.

1B-l-8

Level A0

Level Ax

Arrows Explenation: IIControl

I(Infomation)

Input (Material)- O u t p u t ( I n f o r m a t i o n )- - - - - - ,

- A c t i v i t y

Resouroe~information

flOW;

Figure 13: IDEF 0 - Structure Explanation [6]

GRAI Methodology

The GRAI methodology has been developed bythe GRAI Laboratory 0f Bordeaux University/Francefor the model considers processes. The GRAlconceptual considers a management(decisional) system as a hierarchical structure. Thisdecisional system consists of decision activity centers aspart of the company’s management functions. Decisionframes connect different decision activity centers ofdifferent levels. At least three main decision levels arecommon and the lowest level corresponds to the realtime shop floor level. Besides the decision frames,information links ensure information exchangesbetween decision activity centers. These elements andconcepts will be combined and realized in a "GRAIDecision Activity Grid.” GRAI Grids give ahierarchical representation of the whole structure ofdecision activities in a Production ManagementSystem. In a matrix format functional criteria are used

identify production management functions(columns). Decision time horizons combined withdecision updating cycles criteria are used to identifydecisional levels (rows). This leads to the definition ofdecision time horizons considered for the decision anda revision period (if any) as a time interval for decisionverification. Decision activity centers as building blocks*of the Grid, are the intersections of the functionalcolumns and the time dependent levels.

GRAI-GRIDCODE Character: Roman numbers

Figure 14: GRAI Method - Structure Explanation[V]

Full line arrows within the Grid representdecision links between different decision activitycenters. In a working production management system adecision link leads always from a higher level to alower level or it is used to connect two centers on thesame level, but never from a lower level to a higherlevel. Broken line arrows represent information linksconnecting decision centers without restrictions.

After the most important Decision Activity Gridtwo more Grids can be defined. The intersections offunctions and levels in the Information Grid containthe result of the related decision activity. The ResourceGrid provides information on persons, groups,departments or tools responsible for the relateddecision activity. Therefore it also considersorganizational aspects.

The decision activity centers of the GRAI Gridwill be decomposed into so called GRAI Nets. GRAINets give a more detailed descrition of the variousactivities with the decision activity center. Thisdescription includes the interrelations with otherdecision activity centers. Additionally , because of theimportance of decision activities, variables,. objectives,and rules can be identified and expressed.

Figures 14 and 15 give some more explanationson the GRAI methodology as used within theROCOCO project. Within this context the followingdefinitions should be considered:

1B-1-9

decision frame -

decision objective -

decision variables -

decision rules -

trigger -

request

information -

address

constraint for the followingdecision activity;

aim to be achieved by thedecision activity;

variables allowed to beinfluenced by the decisionactivity;

rules to be followed by thedecision activity obligatory;

information triggeringexecution or decisionactivities;

information given onrequest;

information givenautomatically;

information address -where from/ where to.

Decision Activity

Execution Activity

Additional or alternative elements

Figure 15: GRAI Method - Net ElementsExplanation [6]

Integrated Methodology

The integration of IDEF 0 and GRAImethodologies follows basically the ideas of ESPRITProject No. 418 “Open CAM Systems”. Even hereoperational sub-systems, namely the physicalmanufacturing functions, will be modelled by using theIDEF 0 methodology. The decisional sub-system of themodel will be described by the GRAI approach. Theprinciple connection of the two methods is shown inFigure 16. The control arrow is therefore directly orindirectly (via another IDEF 0 box) coming from theGRAI model. The follow up information to thefollowing IDEF 0 box is described with an outputarrow. The control and output arrows are labeld withaddresses defining the IDEF 0 box/node number orthe GRAI Net Reference number.

1 Textual Explanation sheet

Figure 16: Linkage of GRAI and IDEF 0 [6]

MANAGEMENT FUNCTIONS MODELLING

The GRAI Grid, developed for one-of-a-kindproduction management systems, identifies 8 differentlevels with different management time horizons andperiods (Figure 17). Levels A/B/C are for morestrategic decisions, levels D/E/F have a more tacticalcharacter, and levels G/H should cover the operationaltasks. In this context it 1s difficult to define clear figuresfor horizons and periods for shipbuilding characterizedas an unstable production process with manyunforeseeable events. Therefore thementioned can be considered as typical figures, but a r e

1B-1-10

y - - - - - - - - - - - - - - +

Grid I - Centralized Activities

x =a customery = supplier /subcontractorz = authorities etc.

Figure 17: GRAI Grid Example [6]

Grid IV - Pipe Installation

One-of-a-kind Production Management System Internal System

Es 1 2 3 4 5 6 7 8 9 10 ISFunc

tion External Product Management Material Management Planning Resource Management InternalInformation sale

HIP Purchasing StoreManagement operations Time (Prcduction)

- -a l locate to allocate equlpment

+ - - x i - - + + checkdecide on urgentdesign changes

I I

other work tradestime and attendance

x = customery = supplier I subcontractorz = authorities etc.

. - b descision frame (contains Information relevant for the following decision and a return about the decision making)

(needed to carry out decisions, provided automatically or on request)

Figure 18: GRAI Sub-Grid Example [6]

a c t u a l l o a dof learns

Axx Higher level addresses from/to physical activities.Can be broken further down into all requiredsubaddresses according to IDEFO structure.

Figure 19: GRAI Net Example [6]

1B-1-13

very often disturbed by events. The problem is smallerfor the low level (operational) planning and resourcemanagement activities, i.e. ‘nearer’ to the physicalprocess.

Four main functions have been consideredfurther, broken down into at least 10 subfunctions(columns). Compared with grids developed for seriesproduction, the subfunctions ‘customer contact,’ ‘design.management, ‘operations planning,’ and ‘designpersonnel management’ have to be considered inaddition. Even under the other functions severalactivities have to be mentioned which are of lessimportance to other industries.

The heavily framed activities in the GRAI Gridare specific for different worktrades, and stores. Thedecision activity grid for the pipe installationmanagement process (Figure 18), for example, can beoverlaid. Combined with another grid for a specificstore, a comprehensive grid for one special caseoccurs. The grid structure is completed through ‘shadow’ grids comprising the information generated bythe activity and the resources responsible for theactivity.

The grids should provide a good overview of themost important decision activities. For activity centersof special interest a further detailing through a GRAINet is required. Figure 19 gives an example related toGrid IV (figure 18), activity coordinates G7. Ifrequired, a further break down of the specific decisionor execution activity within this Net might begenerated.

DESIGN AND MANUFACTURING FUNCTIONSMODELLING

The shipbuildingbeen subdivided into five

operational functions havemajor subfunctions. Figure

20 shows a part of the IDEF 0 functions and activitybreakdowns for outfitting sytems manufacturing

functions. In particular the branch for pipe systemrelated functions has been followed here. Furtherdown, three different functions - running the pipestockyard, fabricating pipe systems, installing pipesystems - are distinguished. The “pipe systeminstallation” is shown broken down into two moredetailed levels, allowing the smallest description ofsingle clear-cut activities (modelling elements).

For the breakdown structure reference No. A3232, ‘To install fabricated pi es,’ outlined IDEF 0diagrams: one for material flow and another forinformation flow are shown in figure 21 and 22. Forexample, the presentation of clear- cut activity No. A32322 is given in figure 23.

MODELLING BENEFITS

Beyond the general benefits already mentionedabove under “The need for modelling,” modellingactions produce some other positive side-effects for aCIM approach.

An expert team was chosen to represent thekey-functions of the company and to describe theirworking-system from a top-down view throught theanalysis phase. One effect of the ensuing discussionswas to see the experts learning to recognize and tounderstand some problems of their colleagues.Because of the ‘integration’ aspect of the discussion forsome of the experts, the consequences of omitted workwith no local relevance became more apparent. This isvery important for the generation of the futurereference model from basically the same team. For thevalidation of the results proved d through the top-downapproach, people in the related lower hierarchicalpositions, or those having decision responsibilities forsingle decision activities (GRAI structure), were askedfor their view on the working-system. The effect of theinvolvement of all these people in an analysis anddefinition procedure is to obtain more acceptance ofthe approach itself and to limit distrust of futureimplementation.

A32311 To lssue from staging locationA3231 To Install Units and Components A32312 To Fit units and components

A32313 To Rework (fasten, paint insulate )

Hull Manufacturing Functions

t

A32321 To Issue from staging locationA3232 To lnstall fabricated Pipes A32322 To fit units Fabricated pipes

To Rework (fastern, paint, insulate)

A3233 To Install Key Pips

A3234 To Install scheme pipes

A32331 To Produce cage/wire model A32332 To Stage Model to workshopA32333 To Issue Key pipes from staging locationA32334 To fit Key pipes A32335 To Rework (fasten, paint, Insulate)

A32341 To Messure-A32342 To Fabricate scheme pipe pieces A32343 To Issue scheme pipes from staging location A32344 To Fit scheme -pipesA32345 To Rework (Fasten, paint Insulate)

- - - I F

A32351 To Check completenessA3235 To TestInstalled pipe systems A32352 To (Additoonal work)

A32333 To Prepare Testing A32354 To Test Complete system

Figure 20: IDEF 0 - Subset of an ExemplaryFunction and Activity BreakdownStructure [6]

lB-1-14

USED AT AUTHOR : Juergen Wollert DATE:21/2/90 X WORKING READERPROJECT: ROCOCO-3 REV. : 26/2/90 DRAFT

RECOMMENDED

PUBLICATION A323

staged fabricated

Pipesfabricated

fittingsfitted fabricated pipes

craftsmanship

rework material

DDE:

A3232

craftsmanship WOL9

craftsmanship

TITLE: NUMBER:To Install Fabricated Pipes WOL11

Figure 21:I

IDEF 0 Sheet; Example Material FlowUSED AT AUTHOR : Juergen Wollert DATE:21/2/90 X WORKING READER DATE CONTEXT: WOLll

PROJECT: ROCOCO-3 REV.: 28/2/90 DRAFT c l

COMPANY BIBARECOMMENDED

NOTES : 1 2 3 4 5 6 7 8 9 PUBLICATION A3232

To Extract

w o r k e r i coordinatorassistant'DC-device

worker : coordinator:assistant,:DC-device

worker : assistant,

NODE:

A3232Fl

:DC-device

TITLE: NUMBER:To Install Fabricated Pipes (Information Flow) WOLl2

Example Information Flow

1B-1-15

USED AT AUTHOR : Juergen Wollert DATE:21/2/90 X WORKING R E A D E R

PROJECT: ROCOCO-3 REV.: 27/2/90 DRAFT

CONTEXT: WOLll

COMPANY: BIBA RECOMMENDED

NOTES : 1 2 3 4 5 6 7 8 9 PUBLICATION A3232

jobc a r d s

system fitting installation

P l a n I Plan

II

II

II

I

I

daily work progress---------------------------~

issued fabricated pipesb worked orders

TO FIT FABRICATED PIPES---------------------------+

resource status--------------------------cfittings

b fitted fabricated pipes-

1

worker : coordinator

I: assistant,

craftsmanship

: DC-device tt iI i

i

NODE: TITLE: NUMBER:

A32322 To Fit Fabricated Pipes WOL15

These analysis should be considered by acompany before using external consultant. Theformulation of CIM for every single company must befound internally by the company’s key personnel.External help should be limited to experts for theformal application of the method or for thedevelopment of solutions for specified questions. Theinitiative should always be taken by the problem owner- the company. Then, modelling approaches keep whatthey promise and offer their full benefits.

COMPLEMENTARY ACTIONS

The European shipbuilding industry is realizing,more and more, the challenge of CIM. Through theconsideration of European R + D programs, thecollaboration between major shipyards grows morepermanent.

Figure 23: IDEF 0 Sheet; ExampleModelling Element

lots of different projects have recently been launched.

These projects are too numerous to refer toindividually. However, two representative projectsdealing with modelling in a similar context should bementioned.

ESPRIT Project Nr. 2010 ‘NEUTRABAS(Neutral Product Definition Database for LargeMultifunctional Sytems).

The objective of this project is to developstandardized methods for the storage and exchange ofdata defining shipbuilding and ocean engineering

frproducts. The work is based on standardized interfaceormats such as “Product Data Exchange Specification”

(PDES) and “Standard for the Exchange of ProductModel Data” (STEP) and will define the principles ofapplication oriented reference models relating tocomplex maritime products.

Considering CIM as a threefold problem of:functionality of software (CAD, CAM,PPC etc.);

2. data storage (structure, accessability,data bases etc.); and

3. communication networks andcommunication (user) inter-faces;

It is also expected that the project will provide avaluable contribtuion to international standardizationin the area of data exchange technologies. Contacts toother groups such as the Navy/Industry Digital DataExchange Standards Committee (IDDESC) in theUnited States have been already established.

1B-l-16

ESPRIT Projekt No. 3143 ‘FOF(Factory of the Future Production Theory)

The project is dealing with the develop ment ofnew production philosophies and aworkbench for the

designer'sdevelopment of CIM in production

systems, especially for ‘one-of-a-kind’ products. The

basicproject is running under the framework of ESPRIT research actions. The consortium comprises 7

well established universities and research institutes.The groupideas to to

feels it necessary to look for alternative theory of F.W. Taylor. The shortening of

product lifecycles and the trend towards individualproducts makes it necessary to think about moresuitable solutions for the future compared with thosefor mass production. After other effected products(planea planes, computers), ships play an exemplary role inthis project.To understand the real differences in themanufacturing principles of the effected industry andcompare these with series production (e.g. automotiveindustry), it is a goal of the project to model exemplaryand comparable production activities utilizing the samemethods.

The results, at least generally, promise someanswers for principle questions and as well for theapplicability of CIM within the shipbuilding industry.

CONCLUSIONS

The utilization of diverse modelling methodsand tools to specify the system requirements of CIMfor the shiproduction process facilitates understandinthe manifold and complex interrelations. Method(s)chosen for a ‘common language’ to describe the ‘as is’and the ‘should be’ situation are very necessary. As alanguage understandable for IT experts and matter-experts, modeling can bridge the traditional gapbetween vendors and users. The results are not Justtemporary, but also useful for the documentation ofthe actual situation, for the evaluation and simulationof future concepts, or suitably computerized to managethe implemented CIM elements. This takes the wholemodelling aThe

approach to a new and useful dimension.modeling work leads at least to a living,

functional, and informational, organization andresource model for the company. Even though, thework ISalready of /

just beginning, results are promising andhelp.

Through the acceptance and motivation of themodelling team and through improvements inmodelling techniques and tools, the modellingapproach will definitely lead to benefits for theshipbuilding industry.

ACKNOWLEDGEMENTS

The author wish to express appreciation to allcollegues of the ROCOCO project and the BremenInstitute of Industrial Technology (BIBA) forcomments and contributions to thus paper. Specialthanks go to Jurgen Wollert and Markus Lehne fortheir advice and to Sieglinde Precht without whosehelp this paper would not have been possible.

The conclusions and opinions expressed in thispareflect

Paperer are the author’s own, and do not necessarilyt the opinion of any other source.

BIBLIOGRAPHY

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

Barkmeyer, Edward et. al., “An Architecture forDistributed Data Management in ComputerIntegrated Manufacturing,” National Bureau ofStandards, Gaithersburg,January 1986.

Maryland 20899,

Brodda, J., ‘EDV-unterstiitzteStuicklistenverwaltung in der Einzelfertigung,”Diplomarbeit 1985, Universit Hannover Bunce, P.G.., ‘The Need for Factory Modelling,”CIM Europe Conference Proceedinngs edited byCEC DG XIII, 1987Doumeingts, Guy et al., “Design Methodologyfor AdvancedCommuters in Industry 9 (1987) p. 271-296,

Manufacturing

North Holland"Modelling of Present Situation”, ESPRITProject 2439 (ROCOCO) Report D 11,Consortium Confidential 1989"Preliminary Reference Model”, Report D 13,ESPRIT Project 2439 (ROCOCO), ConsortiumConfidential 1990.‘The Application of CIM to Welded-Fabrication; Industrial Background,” "ESPRI’IProject 595, Technical Report - Vol. 1,1989.“The Application of CIM to Welded-Fabrication; Systems Analysis Methodology,”ESPRIT Project 595; Technical Report - Vol. 2,1989‘The Application of CIM to WeldedFabrication; Definition of User Requirements,”ESPRIT Projekt 595 ESPRIT Project 595 -Technical Report - Vol. 3,1989.“A Primer on Key Concepts and Purpose of theOpen System Architecture for ComputerIntegrated Manufacturing,”ESPRIT Project 688, AMICE & APT, 1987.Kerlen, H., “Methoden und Verfahren derSchiffbaufertigtung - Ziele und Aufgaben fur die90er Jahre,” Jahrbuch STG, Bd. 79,1985.Malloni, J.S.; Nusinow, S.,“An Engineering/Manufacturing EnterpriseIntegrated Information Control System,” ICCAS1985 Conference Proceedings, ShipyardOperation and Ship Design V, EDITED BY P.Banda, C. Kou, G. di Frlippo, North Holland1985.McLean, Charles R..,“Information Archrtecture of the AutomatedManufacturing Research Facility (AMRF),”National Bureau of Standards,Maryland 20899,1986.

Gaithersburg,

Reed, A.W.; Brodda, J.; Marzi, J.,“ROCOCO-Project Overview; in ESPRIT-Section Computer Integrated Manufacturing:Results and Progress of the ESPRIT Projects1989, edited by CEC DG XIII, November 1989.Roboam, M.; Doumeingts, G.; Zanettin, M.,“Study of Manufacturing Systems, Need ofIntegrated Methodology: in EsPrit ‘88. Puttingthe Technology to Use. Part 2, p. 1442 - 1461,edited by CEC DG XIII, Elsevier SciencePublishers B.V., 1988.

1B-1-17

Additional copies of this report can be obtained from theNational Shipbuilding Research and Documentation Center:

http://www.nsnet.com/docctr/

Documentation CenterThe University of MichiganTransportation Research InstituteMarine Systems Division2901 Baxter RoadAnn Arbor, MI 48109-2150

Phone: 734-763-2465Fax: 734-763-4862E-mail: Doc.Center@umich.edu