View metadata, citation and similar papers at to you byCOREprovided by SAM; the Arts et Métiers ParisTech open access repositoryScience Arts & Métiers (SAM)is an open access repository that collects the work of Arts et Métiers ParisTechresearchers and makes it freely available over the web where possible.This is an author-deposited version published in: https://sam.ensam.euHandle ID: . cite this version :Samira SADEGHI, Mohsen SADEGHI, Ali SIADAT, Patrick MARTIN, William ZIZ - Intelligentfeature based resource selection and process planning - Proceedings of the InternationalConference on PLM '10 - Vol. Chapter 3: Process Development, p.133-145 - 2012Any correspondence concerning this service should be sent to the repositoryAdministrator : [email protected]

International Conference on Product Lifecycle Management1Intelligent feature based resource selection andprocess planningSamira Sadeghi1, Mohsen Sadeghi2, AliSiadat1, Patrick Martin1, William Ziz21Laboratoire de Conception, Fabrication et Commande (LCFC) Arts etMetiers ParisTech4 rue Augustin Fresnel, 57078 Metz Cedex 3, FranceTel:0033387375430, Fax : [email protected], Ali.Siadat,Patrick.Martin @ensam.eu2ProctonLabs180 Rue de Vaugirard. F-75015 Paris, FranceTel: [email protected], [email protected]: This paper presents an intelligent knowledge-based integratedmanufacturing system using the STEP feature-based modeling and rule basedintelligent techniques to generate suitable process plans for prismatic parts. Thesystem carries out several stages of process planning, such as identification ofthe pairs of feature/tool that satisfy the required conditions, generation of thepossible process plans from identified tools/machine pairs, and selection of themost interesting process plans considering the economical or timing indicators.The suitable processes plans are selected according to the acceptable range ofquality, time and cost factors. Each process plan is represented in the treeformat by the information items corresponding to their CNC Machine, requiredtools characteristics, times (machining, setup, preparatory) and the requiredmachining sequences. The process simulation module is provided todemonstrate the different sequences of machining. After selection of suitableprocess plan, the G-code language used by CNC machines is generatedautomatically. This approach is validated through a case.Keyword: Process planning, STEP, Knowledge based, integrated design,manufacturing1IntroductionIntegrated manufacturing modelling is the main subject of manufacturing environmentdesign and computerized manufacturing support systems [1], [2]. Principal activitiesinvolved in the integrated manufacturing include product design, process design, suitablemanufacturing process selection and planning [3].The selection of a suitable machining process for a given set of productcharacteristics and requirements is one of the most significant and important issues inproduct development. The early process selection combined with product-processCopyright 2010 Inderscience Enterprises Ltd.

S. Sadeghi, M. Sadeghi, A. Siadat, P. Martin, W. Zizcompatibility analysis determines various parameters to integrate the diverse constraintsimposed by different experts. This analysis is first and foremost based on theexperimental knowledge of the experts. One of the important aspects of knowledge basedintegrated manufacturing design systems is the development of an adequate solution forrepresenting and correlating different types of knowledge. The expert knowledgemodeling facilitates the understanding of the complex interrelationships within integratedproduct-process model in product development.The link between the product, its associated process and resource plan belongs toknowledge formalization. Experts use their knowledge to determine the selection of themanufacturing processes and associated resources or to define the product parametersconsidering the manufacturing and resources constraints.Effective process selection for a given set of product characteristics & requirements isa multi-criterion problem which is strongly influenced by interdependent manufacturingknowledge like product knowledge (product complexity, design requirements, productquality) and resource knowledge (resource availability, characteristic and cost). Based onthe information provided regarding the role and experiences of experts in product-processdesign this knowledge is formalized. The process-resource selection involves two steps:step 1 involves the evaluation of available processes for their technical capability in orderto appropriately respond to design requirements and step 2 ranks the process performancewith regard to its resource consumption economically [4]. Many methods have beenproposed to support these steps. Ashby [5] proposes the selection of manufacturingprocesses considering their compatibility with the geometrical parameter and theresources needed followed by ranking of the relative costs associated. The final choice isperformed with respect to the rating, industrial knowledge, and available manufacturingresources; however, the approach adopted in this research does not target a specificmanufacturing domain.To determine the sequences of specific manufacturing process, Lovatt [6] describesthe framework based on the definition of process plan with different manufacturing tasks(steel cutting, steel heat treatment, casting, etc). The process selection is carried out bythe combination of manufacturing process attributes, material parameters, and productspecifications to satisfy design requirements [7]. The approach proposed by Gupta [8]consists of three phases: (1) identification of the pairs of material/process that satisfy therequired conditions (economical and geometrical specifications), (2) generation of thepossible process plans from identified material/process pairs, and (3) selection of themost interesting process plans considering the economical indicators. In Ishii’swork[9],[10], the process selection is decomposed into two steps: (1) the first step is theidentification of relevant parameters that affect manufacturing processes selection and (2)the second step consists in developing the representation schemas for knowledge usedduring the manufacturing processes selection. This knowledge is used to determinecompatibility between product specifications and process parameters in order to rank thepossible manufacturing processes according to the performance of economical indicatorsmore precisely. Boothroyd [11] propose the selection of the material/process pairs basedon the process capability in terms of geometrical specifications realization and materialconstraint. The final selection is determined by the elimination of the unacceptablematerial/process pairs, considering the related constraints and predefined indicators, suchas the cost.

Intelligent feature based resource selection and process planningIn terms of integrated manufacturing systems, the feature-based approaches havebeen recognized as essential tools to integrate design and automated process planning. Inthis context, Devireddy [12] presented a methodology of integrating the design andplanning of manufacturing by utilizing the concept of feature-based modeling.Khoshnevis [13] described architecture of an integrated process planning system, called3I-PP, which is comprised of three modules: feature completion, process selection, andprocess sequencing. Berger [14] presents an approach for enabling the automaticpreparation of STEP-NC based Workplans with methods known from graph theory.Based on that, it is possible to use algorithms to find the shortest path inside this directedgraph as optimal sequenced solution under given requirements. Finally, thecorresponding NC machining code will be generated and distributed to the machinery.The research results presented in this paper contributes to knowledge-based integratedmanufacturing support by development of a computerized system to process planning,utilizing the concepts of STEP feature-based modeling and rule based intelligenttechniques. The suitable processes plans are selected according to the acceptable range ofquality, time and cost factors. This approach is validated through a PPRint (Product,Process, Resource integrating) tool development and the application of a case study.Table 1, summarizes the characteristics of different manufacturing process selectionapproaches.Table 12Comparison of the approaches “manufacturing processes selection”ApproachChoice ofmanufacturingprocessChoice Non.a.n.a.SadeghiYesYesPPRintYesKnowledge within Process PlanningIntegrated manufacturing is considered as a network of product, process and resourceparameters and expert’s knowledge as a set of constraints [15] applied on the parametricnetwork. In engineering, constraints are formalized as: special design requirements thatmust be satisfied; logic reasoning activity of experts; dependency between designfunctional requirements, design parameters and flow of information between experts tosupport interdisciplinary knowledge integration (figure 1-a).

S. Sadeghi, M. Sadeghi, A. Siadat, P. Martin, W. ZizFigure 1constraint classification in the process planningThe design and related constraint representation provides an integrated productprocess decision making approach to evaluate its consistency and is useful in selecting asuitable manufacturing process and product design parameters [16] (figure 1-c). Our mainobjectives are:To provide an effective data representation model supporting the integration ofdifferent elements (product, processes, and resources) and analysis ofrelationships and related knowledge between themGather and integrate knowledge related to the product definition and themanufacturing process to control the whole production process, allowing us toproduce a better productImproving the exploitation of knowledge capital in new projectsProvide a basis for knowledge capitalization and reuse issued from differentproduct development projectsSo, constraint definition and classification are the key issues for this framework. Weneed to consider practical ways to represent design constraint in process planningenvironment.Constraints in a process planning analysis can be simply classified into constraintsinside each kind of entity model and constraints between different kinds of entity modelsaccording to the product, process and resources view [17] (figure 1-b):Product: Product constraint that represents relationships between entities definedby product model, for example, constraints about dimension, geometry,tolerance, features, etc.Process: Process constraint that represents relationships between entities definedby process model. For example, the dependency constraint between twosequences of machining operationResources: Resources constraint that represent relationships between entitiesdefined by resource model. For example, availability, compatibilityProduct/process constraint: such as constraint about product features propertyand process parameters to realized this features

Intelligent feature based resource selection and process planningProcess/Resources constraint: for example, constraint about choice of tools formachining specific feature according to optimum manner, feature property,quality, cost and machining timeProduct/ Resources constraint: for example the constraint about product propertyand resource characteristic (Product features vs. Machine accessibility)The manufacturing process selection combined with product-process-resourcecompatibility analysis to determine the various parameters and steps as well as thecorrelation of requirement. The purpose is to find a path among several process planningobstacles that could lead to a better solution that satisfies the diverse requirementconsidering constraints imposed by different experts throughout the process planning. Inmanufacturing industries, this analysis is first and foremost based on the experimentalknowledge of the experts.The method we describe in this paper is based on the declaration and checking ofexpert’s knowledge (constraint) in order to check the coherency between product, processand resource elements.With the definition of constraint categories, the system provides the suitableconstraint verification. It allows specifying which constraint representation approach ispossible for each element of data model.We consider following general steps which must be considered to cerate a suitableprocess plan (figure 2):1. The step provides the environment to import the STEP file and to illustrate theproduct information which is necessary to analyze the coherent manufacturingprocess. It contains product constraint which defines the relationship betweenthe elements of product in STEP standard.2. The step determines optimum procedure for machining each feature according tofeature characteristic, part material, tool life, available tool in data base andother constraint.3. The step generate micro process planning according to optimum procedure formachining each feature, spindle speed and feed rate, possible tool direction andorientation for machining and other constraint.4. The step generate macro process by considering the information aboutprecedence machining feature, setup face, machining face, setup time, time oftool moving between each feature and time of tool changing.5. The step presents the selection of coherent CNC machines and the optimumsequences of machining for each CNC machine. The features characteristic andthe tool accessibility knowledge are the necessary elements to select the CNCmachine.6. The step presents the cost model for manufacturing process. It contains differentcosts for different process attributes: physical attributes such as the materials,the tolerance, size range, time, complexity of shape; and economic attributessuch as tooling cost, and rate of production. This module is used to select theacceptable process identifying the cost value specified by the cost model. Theeffect is to eliminate the processes which cannot meet the specifications. Theestimated cost and the time required for the selected process are displayed in theprocess plan, which is the output of this module.7. The step is used to prepare the manufacturing process by generating thecorresponding G-code for the CNC machine. To generate G-code, this modulegets the process plans information as input

S. Sadeghi, M. Sadeghi, A. Siadat, P. Martin, W. Ziz8.The step is used to simulate the machining sequences (process plan). Each stepof simulation demonstrates the machining direction and the tool required torealize the feature.Figure 2 The manufacturing process and resource selection processSTEP file(ARM 224)Design partRead partinformation fromSTEP fileDetermine optimumprocedure for machiningeach featurePossible Tooldirection andorientation formachiningCIllustrate Partinformation in treeview and graphicalformatCDetermine tools formachiningToolsDataMicro process planningCDetermine spindlespeed and feed ratefor machiningMacro process planningCCost computingCNCMachineDataGenerate all possiblealtarnativeSelect suitableCNC machineGenerate G-codeFor CNCMachinCSimulation ofmachining PartThe aim of the proposed approach is facilitated process planning, design andmanufacturing integration and decrease human error in other words allow expressing alltechnical and non-technical constraints through all life cycle phases of an integratedmanufacturing.3Description of the proposed process selection approachThis work focuses on developing a feature-based intelligent process planning. The mainpurposes of the intelligent process planning are: (1) to integrate a standardized featurebased STEP model with process planning utilizing, and (2) developing a knowledgebased and rule-based techniques, and (3) to generate a G-code process plan containing allthe information to setup a CNC machine .

Intelligent feature based resource selection and process planningThe application protocol implemented in this research work is based on STEP AP224which contains explicit high-level part data in terms of feature attributes, tolerances,material specifications. It can be used to represent the information needed to defineproduct data necessary for manufacturing product. The proposed system consists of theSTEP-based feature modeler and intelligent process planning. The figure 3 gives anexample of architecture to realize the process plan (the Holes analysis is represented as anexample). The first step of the system uses STEP features as the basic elements forproduct representation and generates the graphical STEP AP224 based productrepresentation. The product characteristics and the features parameters are represented intree view and graphical format by the feature modeler. This modeler uses the specificalgorithm to convert product information (STEP AP224) into a tree based representationformat to be used as input for process planning. Figure 4 shows an example of featurebased representation of product in the system.This representation provides the part information to execute the following steps:micro process planning to determine the machining procedure for eachfeature according to machining parameters and tools selectionselection of suitable CNC machinemacro process planning to generate the process plansgeneration of G-codes of process plans and simulation possibilityFigure 3 system architectureKnowledge of ToolselectionFeature ParameterHole TypeProductFeaturesProductMicro Process planingDirection of tool formachining featureTool TypeProcedure ofmachining feature forrough and finishingHole DiameterTool DiameterHole HeightTool LengthHole BottomconditionTool MaterialFeed rate and SpindlespeedTool AngleMachining TimeHole.Tool SelectionKnowledge of Toolaccessibility formachining featureHole OriantationTool CostProductMaterialHole PositionProductDimensionMacro Process planingSetup face ,Machining faceSequence ofmachining featuresMaterial CostMachine SelectionMaterialDataSetup Time and Timeof tool movingbetween each featureMachine InformationMachine CostCNCMachineDataKnowledge ofPart accessibilityvs. machineaccessibilityKnowledge of Toolaccessibility formachining featureOptimum procedure forsequence of machining(Moving Time ,Setup ,.)Optimum procedurefor machining feature(Time ,Cost ,Tool life,Quality, .)Knowledge ofprecedenceoperationG-code Generationeach suitable ProcessPlanningSimulation ProcessPlanningToolsDatabase

S. Sadeghi, M. Sadeghi, A. Siadat, P. Martin, W. ZizFigure 4 product characteristics and features representation3.1 micro process planning to determine the machining procedure for eachfeature and parameters and tools selectionThis step selects the machine tools on which the machining operations can beperformed to produce the given feature. This step is implemented using a productinformation, tools data base, tools knowledge as input of the step. Figure 6 presents theuser interface to edit the tool databases. Experts use this interface to complete thedatabase according to available tool information. This step receives information for eachfeature in the part and generates the needed machining operations to realize the feature.The knowledge of tools accessibility for machining feature and tool databases are themain inputs of this step. The outputs of this step are: 1) the direction of tool formachining feature, 2) the procedure of machining feature for rough and finishing, 3) thefeed rate and spindle speed, 4) the machining time, 5) the tool cost.The tools selection is based upon machining feature and its associated machiningoperation. The output of this step contains recommended tools to be used to realize thefeature. These recommended tools can be used as keys to search available tool databasesto find the suitable tool.3.2 selection of suitable CNC machineThe CNC machine selection is realized by considering some constraints about featureorientation and feature position data of each feature which are provided from: 1) designfile, 2) knowledge of tool direction accessibility for machining of each feature and 3)knowledge of accessibility of each type of CNC machine to part. According to theseconstraints and constraints related to part dimension, part weight vs. max table load ofCNC machine, the suitable CNC machine is selected from the available CNC machinesdatabase. Figure 5 shows the user interface to manipulate the CNC machine database.

Intelligent feature based resource selection and process planningFigure 5 user interface of tool databaseFigure 6 user interface of CNC machine database3.3macro process planning to generate the process plansThe process plan generation is a procedure to group the sequence of machining featuresinto process plan. The operation sequence arranges the machining sequences in eachgenerated process plan by considering the information about precedence machining

S. Sadeghi, M. Sadeghi, A. Siadat, P. Martin, W. Zizfeature, setup face, machining face, setup time, time of tool moving between each featureand time of tool changing. In addition, the optimization algorithms are implemented toselect the optimal process according to manufacturing process cost, setup time, toolchanges time among the operations. The system contains a set of rules guiding theprocess planning selection for prismatic parts.As shown in Figure 7, the alternative process plans are proposed for related part. Eachalternative process plan indicates the selected CNC machine and the sequences ofmachining to be executed during setup. The machining features describe what themachine should do and the machining sequences describe how the features should bemanufactured with related sequence tool. The cost presentation allows combination ofseveral cost produced during process plan setup such as material, tools, CNC machineand labor cost to estimate final manufacturing cost. The tool and machine information arealso provided to give the details of tools and machine used during machining sequences.3.4 generation of G-codes of process plans and simulation possibilityAs shown in Figure 8, the specific facility is provided to generate a G-code of physicalprocess plan. This step receives the information related to the part (features, material,etc.) and information produced by process planning (machining operations, tools, processsequences parameters, etc.) to generate a G-code file. Specific module is developed to usethe provided information to simulate the machining sequences in the 3D format.Figure 7 the process plan information detail

Intelligent feature based resource selection and process planningFigure 8 the G-code generation and machining sequences simulation facility4ConclusionIntelligent manufacturing process selection improves the efficiency of manufacturingprocess. In PPRint tool, in the product design stage, the product model is based on theSTEP application protocol, AP 224. A STEP-based feature modeler has been developedto browse the STEP file and to illustrate the product information in the feature tree viewand in the graphical format. In process planning stage, an intelligent rule based processplanning system has been developed to integrate the manufacturing process attributes,material parameters, and product specifications to satisfy design requirements.The approach proposed, consists of different phases, such as identification of the pairsof feature/tool that satisfy the required conditions according to optimum procedure(example: the suitable depth of cut, spindle speed and feed rate that will be chosen duringmachining.), generation of the possible process plans from identified tools/Machine pairs,and selection of the most interesting process plans considering the economical or timingindicators (Example: cost of material, cost of machine, cost of tool, setup time,machining time, tool change time). The suitable manufacturing processes plans areselected according to the acceptable range of quality, time and cost factors. Each processplan is represented in the tree view format by the information items corresponding totheir CNC Machine, required tools characteristics (characteristic, cost), times (machining,setup, preparatory) and the required machining sequences. The process simulationmodule is provided to demonstrate the different sequences of machining to realize thefinal piece. After selection of suitable machining sequences, the G-code language used byCNC machines is generated automatically.In PPRint, the AP224-STEP (machining feature) is used as input, in other wordsaccording to feature based design this format is provided to process planning, but it's notpossible to force designer to design in feature base procedure. In this situation featureextraction is inevitable to capture features parameters for process planning.

S. Sadeghi, M. Sadeghi, A. Siadat, P. Martin, W. ZizReferences1234567891011121314151617Zdrahal Z., Valasek M. and Cermak J. (1999) ‘Selecting Manufacturing Technology: AKnowledge Modelling Approach’, Journal of Intelligent and Robotic Systems, vol. 26, p.405-422.Lee G., Eastman C. M. and Sacks R. (2007) ‘Eliciting information for product modelingusing process modeling’, Data & Knowledge Engineering Vol.62, pp.292–307Tichkiewitch S. and Véron M. (1998) ‘Integration of Manufacturing Processes in Design’,CIRP Annals - Manufacturing Technology, Volume 47, Issue 1, Pages 99-102Esawi A. M. K. and Ashby M. F. (1998) ‘Cost-based ranking for manufacturing processselection’, in Integrated Design and Manufacturing in Mechanical Engineering ‘98,Proceedings of the 2 nd IDMME Conference held in Compiegne, FranceAshby M.F., Brechet Y.J.M and Cebond D. (2004) ‘Selection strategies for materials andprocesses’, Materials and Design, Vol. 25, pp. 51-67Lovatt A.M., Shercliff H.R. (1998) ‘Manufacturing process selection in engineering design’,Materials and Design, Vol. 19, pp. 205-230Zha X.F (2005) ‘A web-based advisory system for process and material selection inconcurrent product design for a manufacturing environment’, International Journal ofAdvanced Manufacturing Technology, Vol. 25, pp. 233-243Gupta M. and Galloway K. (2003) ‘Activity based costing and management and itsimplications on operations management’, Technovation, Vol. 23, n 3, pp. 131-138Ishii K., Lee C.H. and Miller R.A. (1990) ‘Methods for process selection in design’,Proceedings of the ASME Design Theory and Methodology Conference, vol. 27, pp. 105112Ishii K., Krizan S., Lee C.H. and Miller R.A. (1991) ‘HyperQ/Process - An Expert Systemfor Manufacturing Process Selection’, Proceedings of the 6th Int Conf on AI applications inEngineering, pp. 405-422Boothroyd G., DewhurstE P. and Knight W. (1994) ‘Production Design for Manufacture andAssembly’, edited by Marcel Dekker Inc., ISBN 0-8247-9176-2Devireddy CR, Ghosh K (1999) Feature-based modeling and neural networks-based CAPPfor integrated manufacturing. Int J Comput Integ Manuf 12:61–74Khoshnevis B, Sormaz DN, Park JY (1999) An integrated process planning system usingfeature reasoning and space search-based optimization. IIE Trans 31:597–616Berger U. and Kretzschmann R. (2009) ‘A Heuristic STEP-NC Based Process Planning Toolfor Sequencing NC Machining Operations’, Advanced Design and Manufacturing Based onSTEP, Springer Series in Advanced Manufacturing, Publisher: Springer London, pp.49-78L. T. Ly, S. Rinderle and P. Dadam (2008) ‘Integration and verification of semanticconstraints in adaptive process management systems, Data & Knowledge Engineering,Vol.64 pp.3–23Thibault A., Siadat A., Bigot R. and Martin P. (2007) ‘Proposal for Product ProcessIntegration using Classification and Rules’, International IEEE conference on Computer as atool, Warsaw (Poland)Sadeghi M., Noel F. and Hadj-hamou (2007) ‘Toward a framework for effectivecollaborative product development’, in the The Future of Product Development Book(Proceedings of the 17th CIRP Design Conference) published by Springer, pp. 279-289

4 rue Augustin Fresnel, 57078 Metz Cedex 3, France Tel:0033387375430, Fax : 0033387375470 . corresponding NC machining code will be generated and distributed to the machinery. . tool moving between each feature and time of tool changing. 5. The step presen