an interoperable solution for cloud manufacturing

Robotics and Computer-Integrated Manufacturing 29 (2013) 232–247

Contents lists available at SciVerse ScienceDirect

Robotics and Computer-Integrated Manufacturing

0736-58

http://d

Abbre

Broker

Cloud c

CManuf

CNC, Co

CUser, C

plannin

Manufa

Manufa

agent; S

medium

providen Corr

E-m

journal homepage: www.elsevier.com/locate/rcim

An interoperable solution for Cloud manufacturing

Xi Vincent Wang, Xun W. Xu n

Department of Mechanical Engineering Faculty of Engineering, University of Auckland, Auckland 1142, New Zealand

a r t i c l e i n f o

Article history:

Received 29 October 2012

Received in revised form

5 January 2013

Accepted 7 January 2013Available online 26 January 2013

Keywords:

Cloud

Cloud computing

Cloud manufacturing

Service-oriented architecture

STEP

45/$ - see front matter & 2013 Elsevier Ltd. A

x.doi.org/10.1016/j.rcim.2013.01.005

viations: AI, Artificial intelligence; API, Applic

agent; BiA, Billing agent; CAx, Computer-aide

omputing; CDA, Change detection agent; CIA

acturing, Cloud manufacturing; CMService, C

mputer numerical control; CService, Cloud s

loud user; EIA, Enterprise interface agent; ER

g; EU, Enterprise user; GUI, Graphical user in

cturing capability; MCloud, Manufacturing Cl

cturing resource; RRA, Resource recognition

Cloud, Storage cloud; SCM, Smart Cloud man

-sized enterprise; SOA, Service-oriented arch

r; ST, Service template; STEP, STandard for ex

esponding author. Tel.: þ64 93737599; fax:

ail address: [email protected] (X.W. Xu).

a b s t r a c t

Cloud manufacturing is a new concept extending and adopting the concept of Cloud computing for

manufacturing. The aim is to transform manufacturing businesses to a new paradigm in that

manufacturing capabilities and resources are componentized, integrated and optimized globally. This

study presents an interoperable manufacturing perspective based on Cloud manufacturing. A literature

search has been undertaken regarding Cloud architecture and technologies that can assist Cloud

manufacturing. Manufacturing resources and capabilities are discussed in terms of Cloud service.

A service-oriented, interoperable Cloud manufacturing system is proposed. Service methodologies are

developed to support two types of Cloud users, i.e., customer user and enterprise user, along with

standardized data models describing Cloud service and relevant features. Two case studies are

undertaken to evaluate the proposed system. Cloud technology brings into manufacturing industry

with a number of benefits such as openness, cost-efficiency, resource sharing and production

scalability.

& 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Cloud computing (CComputing) is a model for enabling ubi-quitous, convenient and on-demand network access to a sharedpool of configurable computing resources (e.g., networks, servers,storage, applications, and services) that can be rapidly provi-sioned and released with minimal management effort or serviceprovider interactions [1,2]. It provides resources to a user on the‘‘pay-as-you-go’’ basis. There are three common types of CCom-puting structure, i.e., Infrastructure as a Service (IaaS), Platform asa Service (PaaS), and Software as a Service (SaaS). IaaS provides abunch of physical and virtual machines, based on which users areable to install and deploy their own operation systems andworking environments. A PaaS model packages a computingplatform including operating system, programming language

ll rights reserved.

ation protocol interface; BA,

d applications; CComputing,

, Customer interface agent;

loud manufacturing service;

ervice; CU, Customer user;

P, Enterprise resource

terface; MCapability,

oud; MResource,

agent; SA, Supervision

ager; SME, Small and

itecture; SProvider, Service

change of product data

þ64 93737479.

execution environment, database, and web server. A PaaS clientis able to develop and run its applications at the software layer.Finally, SaaS simplifies the utilization of a large amount of soft-ware applications remotely, elastically and seamlessly.

During the past few years, many successful CComputingbusiness cases are found world-widely [3–7]. Among varioustypes of models, the key characteristic of CComputing is that ofpay-as-you-go. In the increasingly globalized manufacturing con-text, customer-oriented manufacturing is a promising approachto improving the service quality and competitiveness, in particu-lar for the small and medium-sized enterprises (SMEs). Thus, anew concept of advanced manufacturing model is proposedworldwide, namely Cloud manufacturing (CManufacturing). Inthe first half of this paper, recent CManufacturing research worksare reviewed, followed by discussions on related works thatsupport building and maintaining a CManufacturing system. Inthe second half, a novel manufacturing platform is proposed toarchive manufacturing interoperability in the Cloud paradigm.

2. CManufacturing—The next generation manufacturingmodel

CManufacturing is a model for enabling ubiquitous, conveni-ent and on-demand network access to a shared pool of configur-able manufacturing resources (e.g., manufacturing software tools,manufacturing equipment, and manufacturing capabilities) thatcan be rapidly provisioned and released with minimal manage-ment effort or service provider interactions [8]. Like CComputing

www.elsevier.com/locate/rcim

www.elsevier.com/locate/rcim

http://dx.doi.org/10.1016/j.rcim.2013.01.005



mailto:[email protected]


X. Vincent Wang, X.W. Xu / Robotics and Computer-Integrated Manufacturing 29 (2013) 232–247 233

concept, manufacturing infrastructure, platform and softwareapplication in CManufacturing can be offered as a service to aCUser. By extending the concept to a broader scope, all theproduction objects and features can be treated as a service, henceeverything-as-a-service (XaaS). The rest of this section discussesthe CManufacturing structure and related technologies.

2.1. State-of-art Cloud manufacturing approaches

Cloud concept presents a promising future for computingbusiness and the same can be said for manufacturing business.Tao et al. [9] proposed a framework of CManufacturing withdiscussions of key advantages and challenges for future CManu-facturing systems. CManufacturing is described as a computingand service-oriented manufacturing model developed fromexisting advanced manufacturing models (e.g., Application serviceprovider, Agile Manufacturing, Networked Manufacturing, andManufacturing Grid), enterprise information technologiesunder the support of cloud computing, Internet of things, virtua-lization and service-oriented technologies, and advanced comput-ing technologies. It is predicted that a CManufacturing systemwould reduce the cost and increase the utilization rate ofresources. Li et al. [10] proposed a service-oriented networkedmanufacturing model. The paper also discussed a number ofmethods to support the model. Intelligent agent, Product LifecycleManagement (PLM), resource modelling and evaluating technol-ogies are considered as the supporting technologies for Cloudarchitecture.

A Cloud-based manufacturing research project [11] waslaunched in Europe in 2010, sponsored by the European Commis-sion. The goal of this project is to provide users with the ability toutilize the manufacturing capabilities of configurable and virtua-lized production networks. CManufacturing is described as aservice-oriented IT environment as a basis for the next level ofmanufacturing networks by enabling production-related inter-enterprise integration down to shop floor level [12]. A set ofsoftware-as-a-service applications have been developed. In theproposed system, customized production of technologically com-plex products is enabled by dynamically configuring a manufac-turing supply chain [13,14]. It is believed that the development ofa front-end system with a next level of integration to a Cloud-based manufacturing infrastructure is able to better supporton-demand manufacture of customized products.

To facilitate a CManufacturing environment, existing resourcesneed to be scaled, modelled and adapted into the Cloud. Wuand Yang [15] proposed a method to describe and scale manu-facturing resources in a Cloud. Hu et al. [16] analysed thefactors that affect the classification of virtual resources in CMa-nufacturing. Examples are introduced to validate the effect ofthese factors to task assignment. Luo et al. [17] discussed aCManufacturing system from the viewpoint of network, functionand running. A multi-dimensional information model was pro-posed to describe manufacturing abilities [18]. This knowledge-based data model helps provide a user with manufacturingservices via network.

To control and manage the flexibility of the resource servicecomposition in CManufacturing, Zhang et al. [19] proposedarchitecture considering major uncertain dynamic changing fac-tors in the life-cycle of a resource service. Multi-agent is provedto be an effective tool in solving problems through sharingknowledge during the implementation of CManufacturing [20].An Agent-based Mechanism provides flexible and effective sharingand utilizing of elastic resources.

After resource modelling, the next challenge is resourceintegration. Fan and Xiao [21] proposed integrated architectureof CManufacturing based on a federation principle. Federation

integration rules are applied before resources are connected tothe system. Thus, joining or exiting of a resource would not affectoperation of the whole Cloud environment. To maintain theCManufacturing resources, an Optimal Allocation of ComputingResources (OACR) system was proposed [22]. In OACR, improvedNiche immune algorithm is introduced to solve the resourcescheduling problem in a grid system or CComputing system,associated with the Niche strategy.

2.2. Supporting technologies for Cloud manufacturing

Although CManufacturing is still a relatively new concept, itdraws upon technologies such as virtual enterprise, distributedand collaborative manufacturing systems. Xu [8] reviewed thesystematic requirements of CManufacturing systems. Advancedmanufacturing technologies are discussed to fulfil these specifica-tions and support a CManufacturing environment. Research con-tributions are reviewed regarding collaborative manufacturingand interoperable systems [23]. Manufacturing systems arere-evaluated from the Cloud perspective, e.g., IaaS and SaaS. Inaddition, ERP (Enterprise Resource Planning), SOA (Service-Oriented Architecture), and modelling systems are also relevantto the concept of CManufacturing.

After the manufacturing activities are properly modelled,the next step is to integrate their operational processes. Torepresent a business in a manufacturing Cloud (MCloud), thefirst step is to understand and model an enterprise. ERP systemshave been studied extensively [24–28], including the inter-organization performance [29] and the behaviour throughoutthe supply chain with multiple stakeholders [30]. With the helpof ERPs, inter-organization behaviours/reactions can be modelledand mapped in a standardized manner as neutral APIs (Applica-tion Protocol Interfaces). Based on these APIs, MCloud can beestablished via integrating these reactions in standardizedsemantics, without changing the organizational structure of anenterprise.

Papazoglou and van den Heuvel [31] proposed a frameworknamed Enterprise Service Bus. It is an integration platform thatutilizes web services standards to support SOA applicationswithin an enterprise. The extended SOA system can be furtheradopted by CManufacturing to enable capabilities such as serviceorchestration, ‘‘intelligent’’ routing, provisioning, integrity andsecurity of message as well as service management. In an SOAsystem, business procedures can be modelled and componentizedto support seamless business integration [32]. Schmidt et al. [33]proposed architecture declaring clear definitions of service cap-ability and requirements in a service-oriented context. Modelshave been proposed to evaluate the quality/feedback in thebusiness-to-business context [34].

It has been suggested that a commonly used data model/schema should be utilized for a wide range of products [35].Data management should be encapsulated by schema and manip-ulation rules in a data model. In the perspective of CManufactur-ing, international standards, e.g., STEP/STEP-NC have a roleto play in ensuring product data interoperability. STEP (theStandard for Exchange of Product data [36]) is one of suchstandards, providing mechanisms for describing product informa-tion throughout the lifecycle. Different Application Protocolshave been developed for different applications/domains. As anextension of STEP, STEP-NC [37] is developed to support CNC(Computer Numerical Control) manufacturing. Compared withprevious standards, these data models offer a set of effective toolsfor interoperability solutions in the computer-aided manufacturingcontext [38].

X. Vincent Wang, X.W. Xu / Robotics and Computer-Integrated Manufacturing 29 (2013) 232–247234

2.3. Recap

CManufacturing is not just the implementation of CComputingin manufacturing. Manufacturing enterprises and relatedresource/capability need to be described, componentized, virtua-lized and integrated in a MCloud. Some existing research workprovides some enabling tools to the CManufacturing concept. Thissaid, there is still a lack of Cloud solution for the entire manu-facturing supply chain. It is necessary to implement a supervisionmechanism to organize and control Cloud services (CService) atthe upper level. Moreover, an interoperable environment is alsoneeded to integrate current and future manufacturing resources.The next section describes a proposed service-oriented CManu-facturing system named Interoperable Cloud-based Manufactur-ing System (ICMS).

3. Interoperable Cloud-based manufacturing system

Nowadays, a manufacturing enterprise would not survivewithout computer-aided applications (CAx) technologies. Deploy-ing CAx software on the Cloud improves the performance in termsof flexibility, extendibility, integrity and easy/unlimited datastorage. With a Cloud structure, software is easily maintainedand utilized on a Cloud server. Version updating, maintaining andintegrating is remotely done by the Cloud provider, whichreplaces periodic services by on-site maintenance specialists.Thus, the cost of IT infrastructure is cut down via reducedmanagement and maintenance effort. Additionally, thanks to thepay-as-you-go basis of a CService, the cost of expensive applica-tions is spread over multiple CUsers. Costly but rarely usedsoftware can be priced by the amount of usage.

CManufacturing faces a tougher challenge than implementingmanufacturing-related software in CComputing. Unlike softwareprogramme and IT infrastructure, physical machines, monitors,and facilities cannot be readily deployed on the Cloud. There isalso a need to understand the intermediate processes from rawmaterial to finished products.

3.1. Manufacturing capability and manufacturing resource

Zhang et al. [39] identified manufacturing ability as a kind ofresource. In practice, the main reason for acquiring a manufactur-ing facility is the functionality of the equipment but not theequipment itself. It is therefore necessary to recognize a Cloudresource, capability, and service at different levels. In the Cloudbackground, the definitions of a resource, capability and serviceare given below,

�
Manufacturing resource (MResource): material and nonmater-ial manufacturing supplies including equipment, machine,device and intelligent properties. � Manufacturing capability (MCapability): ability of transform-
ing one form into another in manufacturing domain. It isrealized via related MResources.
� Cloud manufacturing service (CMService): Self-contained, con-
figurable and on-demand manufacturing service package tofulfil user’s original needs. A CMService can be random, short-term, long-term, or strategic.

The containment relationships of MResource, MCapability andCMService can be summarized as shown in Fig. 1. MResources arecontained within MCapability as one of the essential require-ments, since MCapability is realized and implemented via MRe-source. MCapabilities are re-packaged and deployed in the

MCloud as CMService as a convenient feature that can be rapidlyprovisioned and released by a CUser.

A CManufacturing system encapsulates and implementsMCapability in the Cloud as CMService packages. Manufacturingcapability is composed of design, production, experimentation,management, and communication capability.

�
Design capability (DC) refers to domain-specific design knowl-edge, expertise of the organization and past experience fromprevious design activities. � Production capability (PC) relies on the speed and quality of
creating an output, i.e., product or service, to fulfil aproduction order.
� Experimentation capability (EC) entails the experimentation
knowledge and specialists.
� Management capability (MC) includes planning, organizing, staff-
ing, leading and controlling of an organization. It relies on theability of the operational business and organizational activities.
� Communication capability (CC) refers to the data exchange-
ability between applications/devices. It includes data trans-portation, speed, storage, conversion and QoS.

From the resource’s perspective, each kind of manufacturingcapability requires support from the related MResource(s). Foreach type of MCapability, its related MResource(s) comes in twoforms, soft resources and hard resources. The soft resourcesinclude:

�
Software: software applications throughout the product life-cycle including design, analysis, simulation, process planning,etc. � Knowledge: experience and know-how needed to complete a
production task, i.e., engineering knowledge, product models,standards, evaluation procedures and results, customer feed-back, etc.
� Skill: expertise in performing a specific manufacturing task. � Personnel: human resource engaged in the manufacturing
process, i.e., designers, operators, managers, technicians, pro-ject teams, customer service, etc.
� Experience: performance, quality, client evaluation etc. � Business Network: business relationships and business oppor-
tunity networks that exist in an enterprise.

Hard resources contain:

�
Manufacturing equipment: facilities needed for completing amanufacturing task, e.g., machine tools, cutters, test andmonitoring equipment and other fabrication tools. � Monitoring/control resource: devices used to identify and
control other manufacturing resource, for instance, RFID(Radio-Frequency IDentification), WSN (Wireless Sensor Net-work), virtual managers and remote controllers.
� Computational resource: computing devices to support pro-
duction process, e.g., servers, computers, storage media, con-trol devices, etc.
� Materials: inputs and outputs in a production system, e.g., raw
material, product-in-progress, finished product, power, water,lubricants, etc.
� Storage: automated storage and retrieval systems, logic con-
trollers, location of warehouses, volume capacity and sche-dule/optimization methods.
� Transportation: movement of manufacturing inputs/outputs
from one location to another. It includes the modes of

Fig. 1. MCapability and MResource.


transport, e.g., air, rail, road, water, cable, pipeline and space,and the related price, and time taken.

To formulate MCapability, a MCapability Description Model(MCDM) as a 4-Tuple is proposed,

MCapability¼ fDCðRSoftDC, RHardDCÞ,

ECðRSoftEC, RHardECÞ, PCðRSoftPC, RHardPCÞ,

MCðRSoftMC, RHardMCÞ, CCðRSoftCC, RHardCCÞg ð1Þ

where, R is the MResource, all the resources required to carry outthe Task, including hard resource RHard and soft resource RSoft.

High-performance service needs sufficient resource and suita-ble methodology to exploit it. Hence, an effective MCapability iscontributed by the domain-specific ability and its relatedresource. MCDM includes the capability of both an individualenterprise and an alliance made up of multiple participants. Thismeans a MCapability meeting a CUser’s need could be providedby a single Service Provider (SProvider) or a union of them. Acomprehensive Cloud solution is required to take care of all thecapabilities and resources mentioned above and provide anoptimal solution. Eventually, identified MCapabilities are pack-aged as CMServices and deployed in the MCloud. During theconversion from current manufacturing status into CManufactur-ing, existing capabilities and resources should be integrated andutilized in the CManufacturing environment. Thus, an interoper-able, service-oriented CManufacturing system can be realized.

3.2. ICMS architecture

As mentioned above, Cloud technology provides an opportu-nity to re-shape manufacturing business, in particular SMEs.Combined with SOA, it is capable of creating new economicgrowth for customized production or one-of-a-kind production(OKP) businesses. Specialized and customized demands can bebetter served due to the flexible and fast-reaction nature of aCManufacturing system. Compared with the business-to-business(B2B), business-to-consumer (B2C) models, an X2C (Everything-

to-Cloud) model is proposed. The preliminary concept of ICMS hasbeen reported in [23]. As public Cloud infrastructure, ICMSconsists of three layers, i.e., Smart Cloud Manager (SCM), UserCloud (UCloud), and MCloud (Fig. 2).

3.2.1. Customer and enterprise user

At the UC layer, the CMService consumer is divided into twocategories: customer user (CU) and Enterprise User (EU). ICMStakes care of traditional manufacturing tasks for CUs as well ascollaborative production requests from multiple organizations(Fig. 3). By combining the Consumer-to-Cloud and Business-to-Cloud model, ICMS provides an X2C structure from the industrialcontext.

CU is defined as a customer or organization with the request ofa self-contained production task. Assisted by the Customer Inter-face Agent (CIA) of SCM, the manufacturing request of a CU isanalysed and located by SCM, and provided by the MCloud. Thus,it forms a Request-Find-Provide service chain. Original user’srequests are taken care of by SCM. SCM searches for potentialsolutions and feeds back the results to the user. The user is able tooptimize the solution based on his/her original needs and finalizethe service request. ICMS provides a user with a big range offlexible manufacturing capabilities. Customized and originalrequirements can be realized easily, compared with the tradi-tional manufacturing practice. For industry, it offers new oppor-tunities especially for the OKP enterprises and SMEs. Theenterprises are loosely integrated in MCloud as ICMS SProviders.MCapabilities and business opportunities are integrated andbroadcasted in a larger resource pool, which enhances thecompetiveness of the entire consortium. Thus, more manufactur-ing objects can be achieved with minimum additional investmentand effort.

Besides CUs, ICMS also takes care of organizations/enterprises(EU) who are seeking additional MCapabilities and supports. Inpractice, customers occasionally come to a manufacturing enter-prise requiring products or capability that the enterprise by itself

Smart CloudManager

Supervision Agent

Broker Agent

ProjectDatabase

Service ApplicationCloud

ProductDatabase

CustomerUser

ManufacturingCapability

Layer

VirtualServiceLayer

ApplicationLayer

(User Domain)

Enterprise Interface Agent

Storage Cloud

Central Server

Backup Database Segmentation

ManufacturingCloud

Enterprise User

Customer Interface Agent

Visual Drag-and-Drop InterfaceUser Cloud

Smart CloudManager

Resource RecognitionAgent

Change DetectionAgent

Billing Agent

Data PacketGenerator

STEP- NC Adapter

Simulation

NCK

CAPP Module 1

CAM Module 1

CNC Module 1

CNC Module 2

ResourceDatabase

Service Provider 2Service Provider 1 Service Provider 3 Service Provider 4

Physical ResourceDomain

KnowledgeDatabase

ClientDatabase

ProviderDatabase

Firewall

Fig. 2. ICMS architecture.


cannot fulfil. With the help of the enterprise interface agent (EIA),an EU can search for qualified SProviders who are able to ‘‘fill inthe gap’’. The EU is able to recognize related MResources andallocate the temporary partner(s) for the task. In this case, theoriginal EU plays a role of the ‘‘leading company’’ in this virtualorganization. The leading company is in charge of interacting withthe customer, and collaborating with other participants as acoordinator. From the ICMS perspective, the leading company isconsidered as the EU, who will be assisted by the SCM module.This way, the CUsers are able to accomplish bigger and moredemanding production tasks that are otherwise not possible by asingle enterprise. As a matter of fact, the partner network of acompany is made no bounds.

3.2.2. Smart Cloud manager

Intelligent agent technology is capable of supporting manu-facturing procedures/decisions [40–42]. The SCM module is con-structed by intelligent agents. In an ideal system, user shouldhave full confidence of the system’s intelligence. The interactionbetween a human being and system intelligence should beminimized as long as the service request is well-defined by theuser. Intelligence kernel is capable of optimizing and executingthe task with the preference variables from the user. However,manufacturing decisions are difficult to make due to complexityof manufacturing processes, variety of machines/devices, anduncertainty of resources status. When multiple resources andvariables are involved, it is even harder to predict a reliable and

optimum service solution for the user. Thus, to fully utilize AI andhuman expertise/knowledge, a decision-making model is pro-posed. SCM works in a neutral manner and consists of EIA, CIA,Broker Agent (BA), Supervision Agent (SA), and Firewall Module(Fig. 4). EIA works with EUs, and CIA handles requests from CUs.Although the GUIs (graphical user interface) and algorithms of EIAand CIA are different, the service procedures are almost the same(from the SCM perspective). After the user’s request is collectedby the interface agent (IA), BA communicates with the providerdatabase and maps the requirement to the available CMServices.As long as the user modifies and confirms the service package, anICMS Service Template (ST) is generated and delivered to theSCM. Based on ST, supervision agent starts up and works with theService Application Cloud (SACloud). Specific CMServices areorganized and launched to meet the users’ expects. The finalservice output, which can be product, computing data or technicaldocument, is sent to the user. After the feedback/evaluationdocument is finished by the user, the CMService is terminated.

As the supervisor or brain of ICMS, SCM analyses and controlsthe CMServices to fulfil the user’s demand. Inside SCM, theinteractions among IA, BA, and SA are summarized in Fig. 5. Aftera user’s request is collected by the IA, the details are convertedinto a standard format. Based on these details an internal requestdocument is generated and sent to BA. According to the requestdocument, BA searches in the MRsource database for potentialsolutions. Afterward an initial ST file is created and sent back tothe user. Cloud consumer is able to view all these solutions alongwith the suggestions from SCM. Based on factors such as cost,

Enterprise User (EU)

Enterprise 1 Enterprise 2 Enterprise 3

Enterprise 4 Enterprise 5 Enterprise 6

D S M G WRequest:

D M GService: T

S MService: T DService: S T WService: M

D M GService: T WService: MService: S

S? W? Better T?

Solution: EU + E2 +E5

ManufacturingCloud

Customer

Customer User (CU)

D S M G WRequest:

Solution: E1 + E2+ E5

D – DesignM – MillingT – TurningW – WeldingG – GrindingS – SimulationO – Optimization

Smart CloudManager (AI)

CIA EIA

Smart CloudManager (AI)

CIA EIA

Fig. 3. Customer user and enterprise user.


quality, functionality etc., SCM recommendation is visible to theuser in different levels of details. If the Cloud customer is notsatisfied with any of the suggestions provided, he/she is able tomodify the ST.

At this stage, the Cloud customer is able to either optimize theST via BA intelligence or do it manually. If the customer prefers toutilize AI continuously, CUser is requested to modify his/her

original searching request by providing more details or to modifytechnical variables. Then, the altered request condition is sentback to BA, who will process one more round of analysis andservice detection. On the other hand, if the Cloud customerchooses to improve the ST manually, he/she can work on it viaGUI and allocate a preferred provider. This way, both of expertknowledge and optimization are utilized in SCM.

Application ModuleCloud

User ICMS

OR

Enterprise User

Customer User

1. UserRequest

2. Feedback

3. Modify

4. Update

5. Confirm

6. Execute

1. EnterpriseRequest

Products

Smart CloudManager

Supervision AgentBroker Agent


Firewall


ProviderDatabase

Smart CloudManager

Supervision AgentBroker Agent


Firewall


7. Feedback& Evaluation

8. ServiceCompleted

Fig. 4. Cloud service procedure.


As long as the user confirms ST, the specific Cloud Services arelaunched by the SA. SA is responsible for monitoring and con-trolling all the activities of the Cloud service modules. By markingand manipulating the event/data flow of all the applicationmodules, the ST is executed accurately as it is defined.

3.2.3. Manufacturing Cloud: Provider and service database

At the manufacturing capability layer, MCapabilities are inte-grated as self-contained service modules in the SACloud. Theoperational processes throughout the supply chain stay in form ofCMService applications at this layer. By controlling the serviceinput and output, CMServices are shared and published in thehigh-performance resource pool. The plug-and-play ability ofservice applications enables flexibility and adaptability to copewith uncertain and changing manufacturing market.

In Storage Cloud (SCloud), database maintains the product/project data as well as the information of the MResources meshedin SACloud. To model and recognize these application modules,three smart agents are developed, i.e., Resource Recognition Agent(RRA), Change Detection Agent (CDA) and Billing Agent (BiA). RRAis responsible for identifying newly-published capability andtermination of existing ones. Since MCapabilities are looselymerged in the SACloud, the performance of the entire MCloud isnot affected by binding or detaching an individual CMService.CDA is in charge of detecting and updating resource changes, suchas its availability, price adjustment, facility maintaining, etc. Thus,the up-to-date data of resource can be supplied while SCM issearching for applications for the CUser. BiA works with StorageCloud and SCM directly. When ST is generated, BA provides theservice quote based on the pre-defined service description in SC.Thus, real-time information exchanging between MCapabilitiesand MCloud is enabled.

To clearly describe and present CMService, CProvider andCUser’s requests, efficient data models are needed. A number ofmodelling languages can be used, e.g., Web Services Description

Language (WSDL) for web-service description, Web OntologyLanguage (OWL) for knowledge representations, Web ServicesBusiness Process Execution Language (WS-BPEL) for executablebusiness processes with web-services, and EXPRESS. Amongthem, EXPRESS is chosen since it provides more robust modellingmethods. EXPRESS is a standardized data modelling language forproduct data which is formalized in an ISO standard [43]. As thegraphical notation of EXPRESS, Provider & Service models inEXPRESS-G provide the portability with standard data modelssuch as STEP and STEP-NC.

As shown in Fig. 6, the enterprise which provides CMServices isdefined as a SProvider. Therefore, a manufacturing enterprise can bedescribed in Cloud terminology, as provider profile and serviceproperties. The provider specifications describe the information ofthe organization, while service specifications present the MCapabilityin terms of service that it provides. Note that one company has aunique Enterprise Entity, while its Service entities can be multiple.Hence, the organization consistency and service variety are main-tained concurrently. Entity Enterprise outlines the properties of aCManufacturing via entities Provider_ID, Company_Name, Provider_-Size, Provider_Capability, Provider_Location, Provider_Contact,Prior_Experience, Provider_Evaluation and Provider_Description.

Entity Provider_ID provides a unique identifier in the MCloudfor a SProvider. Based on its Provider_ID, all the CMServices froma provider and its related service history can easily be traced.

Entity Provider_Capabilities describes the MCapability of aSProvider via sub-entities Design_Capability, Experimentation_-Capability, Production_Capability, Management_Capability, andManufacturing_Resource, which are compliant with the afore-mentioned MCDM model. Entity Hard_Resource and Entity Soft_-Resource describe the MResources that support a specificMCapability. These entities can be connected to a standardizeddata model directly, for example ISO14949-201 for machine tools,and ISO10303-45 for material and engineering properties. Hence,the MCapability of a SProvider is described in an explicit andscalable data model.

Customer Request

GUI GeneratingRequest Document

(RD)

Smart ManagerAnalysing RD

Smart ManagerGenerating Service

Document (ST)

STfeedbacktoCustomer

Customerapproving ST ?

Cloud ServiceLaunched based on

ST

Yes

Service Outputting

Customer ModifyingRequest Details

No

CustomerConfirmed final ST

AI Optimization ?

Yes

CustomerSuggesting ST

Manually

No

Broker AgentSearching/Finding

Solutions

Interface Agent

SupervisionAgent

Smart CloudManager

Broker Agent

Fig. 5. Logic flow within smart Cloud manager.


Cloud_Service

its_provider

Provider_CapabilityEnterprise

Provider_ID

its_identifier

its_mcapability

Provider_Location

Provider_Contact

its_region

its_name

Design_Capability

Manufacturing_Resource

its_dc

its_mr

Staff

Address

Phone

E-mail

contactor

address

electronic_mail

(ABS)Cloud_Service_Speci

fication

its_service_specification

(ABS)Cloud_Service

_Provider

its_provider_specification

its_service_boundaryL[0:?]

Company_Name

its_contact

Provider_Size

its_size

Provider_Evaluation

Prior_Experience

evaluation_record

privider_history

Production_Capability

its_pc

Hard_Resource

Soft_Resource

its_hr

its_sr

Experimentation_Capability

Management_Capability

its_ec

its_mc

General_Description

general_context

Communiation_Capability

its_cc

phone_number

Fig. 6. Cloud service provider model in EXPRESS-G.


Entity Prior_Experience records the service history of oneSProvider which is visible to the Cloud administrator and provideritself, but not entirely to the users. Entity Provider_Evaluationdocuments the feedback from these service consumers. Based onthese two entities, the performance of the service experiences aremodelled explicitly.

As the second category of enterprise attributes, the recognitionof the specific CMService is modelled via Entity Cloud_Service_-Specification and its entities, i.e., Service_ID, Service_Cost, Price,Time, Shipping_Price, Shipping_Time, Service_Status, Service_Do-cument, Data_Object, Pre_Condition, Availability, Resource, Qual-ity_Evaluation, Technical_Support_Capability, Warranties andService_Description (Fig. 7).

Entity Service_Cost documents the value of CMService in themonetary form. This entity provides an explicit model of the valuethat has been used to accomplish a service object. Service_Cost isonly visible for SProviders to understand their MCapabilityinternally, and to make reasonable price for external CUsers.Service_Cost is described with the help of entities Cost_Per_Unitand Cost_Quantity. The Service_Cost model structure is inheritedby four subtypes, which divides the service cost into fourcategories, Cost_Material, Cost_Machining, Cost_Labour, andCost_Management.

Entity Service_Status contains the information about the run-ning stages after a CMService is launched. Stages of implementa-tion are described via entities Initial, Ready, Running, Skipped,Completed, Overload, Paused, and Cancelled.

Entity ST_Document keeps the path and version of a ST file asabove-mentioned. When a CUser is working on a ST, all versionsof the ST are recorded by this entity. Thus, the service/modifica-tion history is maintained.

Entity Data_Object records the technical document(s) relatedto the CMService. The optional attribute of this entity is EntityProject, which is compliant with the top level of a neutral dataformat defined in ISO10303 [44].

Entity Pre_Condition defines the requirements prior to thestart of CMService. Limitations or preparations of the serviceinput are recorded and published, e.g., limits of size, materialpreparation, heat treatment and so forth.

Entity Availability represents the availability and workingcondition of a CMService. This entity is dynamically updated byCDA. With the help of CDA, CUser is informed by the trust-worthysituation of availability without major delays. CUsers are able toselect the available CMService only, or queue in the list waitingfor the preferred package till it is ready-to-be-used. The avail-ability information is further described by its attributes, i.e.,Module_Completion, Function_Fitness and Security.

Entity Resource defines the manufacturing resource that isrequired for a specific service. Its structural attributes (Hard_-Resouce and Soft_Resource) are compliant with the resourcerepresentation of Entity Enterprise. Thus, the resource specifica-tions, from both service point of view and enterprise point ofview, are shaped and integrated in SCloud.

To describe the user’s query of a CMService, Cloud_Service_Request model is used (Fig. 8). Cloud_Service_Request is compli-ant with the SProvider and Cloud_Service_Template data struc-ture. As a bridge between user’s original demand and CMServicein the MCloud, it provides a neutral and standardized methodol-ogy to document the query. Via GUI, the service description isarranged in the structured statement and transmitted to the SCM.Based on this piece of data, SCM is able to suggest the solutionfrom the resource pool based on the terms and mapping

Pre_Condition

its_precondition

(ABS)Cloud_Service_Template

Service_IDservice_id

Price

Time

Shipping_Price

Shipping_Time

price_of_service

length_of_service

length_of_freight

price_of_freight

ST_Document

Data_Object

its_st_document

its_product_document

Quality_Evaluation

quality_statement

Project

Service_Status

status_of_service

Initial

Ready

Running

Skipped

Completed

Service_Description

general_description

Resource

Overload

Paused

Cancelled

Service_Performance_Report

evaluation_document

Availability

Module_Completion

Function_Fitness

Security

its_completion

its_status

security_level

its_availability

Technical_Support_Capability

Warranties

technical_support

its_warranties

Hard_Resource

Soft_Resource

related_resource

Service_CostCost_Per_Unit

Cost_Quantity

unit_price

amount_of_serviceCost_Material

Cost_Machining

Cost_Labour

Cost_Management

cost_definition

related_resource

Fig. 7. Cloud service model.


preference. The request data is shaped via entities such asRequest_ID, CUser, Request_Status, Service_Type, Data_Object,Service_Time, Service_Price, Preference, Request_Description,Service_Document, and optional entities Preferred_Resource,Preferred_Provider, Preferred_Region, Quantity_Of_Service_Output,and Keywords.

Entity Request_ID gives a unique serial number for a CUser’srequest. When a new query case is created, a permanent Reques-t_ID is assigned. Users are able to resume, modify, and review therequest case. Additionally, all the related Cloud behaviour andhistory are traceable based on a Request_ID.

Entity Request_Status maintains the operational conditionof a request. The variables of a process status are Created,Under_Evaluation, Approved, Running, Waiting, Terminated, andFinished.

Optional Entity Preferred_Resource keeps the user’s predilec-tion of resource and facility, for example specific machine tools,testing method or design software. The structure of this entity iscompliant with the Hard_Resource and Soft_Resource entities ofSProvider model aforementioned. Thus, the user’s request can bedirectly connected to the MCapabilities in the Cloud.

Thanks to the service request, CMService and SProvider mod-els, the data is modelled from the initialization to implementationstage in the SCloud. Information packages can be submitted,retrieved, and maintained over the Internet regardless the loca-tions of the central database and server. For data storage queries,customer’s private data is not maintained in the SCloud directly.In the background, data centers are hosted by third parties thatare integrated as SProviders in the MCloud. Thus, the storagetask is integrated as one of the CMServices in the virtualizedservice pools.

3.3. Recap

To recap, ICMS provides a flexible and distributed environmentfor shared MCapabilities. In particular, it offers a number ofbenefits:

�
Data Interoperability: manufacturing business is commonlytroubled by data interoperability issues. CAx applications arewidely utilized throughout the production stages. However,these applications are provided by multiple providers usingdifferent programme languages and document formats, lead-ing to a heterogeneous data environment. Software tools usingdifferent kernels are difficult to communicate with each other.Data loss and errors often occur during format conversions. Byusing ICMS, standardized (STEP-based) communication meth-odologies have been deployed to support collaborative inter-actions in the Cloud environment. Moreover, ICMS offersexplicit specifications of manufacturing resources in theMCloud. Detailed descriptions, e.g., input and output formatrequirements, are visible to all the CUsers. Interoperableproblems can be easily identified and avoided. Users are ableto choose the SPs that can smoothly communicate with eachother, or alternately allocate reliable data conversion servicebefore-hand as one of the CMServices. Therefore, interoper-ability is achieved even before a CMService is launched. � Globalization/Sub-Contracting: with the help of Internet of
Things, manufacturing services/capabilities are virtualized inthe MCloud. Compared with web-based manufacturing, ICMSprovides a more distributed and flexible environment whichknocks down the boundaries between organizations/enterprises.

Request_Status

(ABS)Cloud_Service_Request

Request_IDi t s_i d

Service_TypeService_Subtype

Service_time

Service_Price

its_price

length_of_service

ST_Document

Data_Object

its_st

Project

Preference

Time_Upper_Bound

Time_Lower_Bound

Price_Upper_Bound

Price_Lower_Bound

Maximum_QoS

its_limit

its_limit

Request_Description

general_specification

Minimum_Cost

Availability

Keywordsits_keywords

Created

Under_Evaluation

Approved

Running

Finished

its_status

CUserUser_Region

Client_ID

its_owner

Preferred_Resourceprior_resource

Preferred_Region

Hard_Resource

Soft_Resource

i t s_resource

prior_region

Preferred_Providerprior_provider

Quantity_Of_Service_Outputquantity _of_service

Type_Of_User

Waiting

Terminated

its_product_document

its_service_type

service_statusoptimizatio_npreference

Cost_Effectiveness

Energy_Efficiency

Fig. 8. Service request model.


It is easier to find business partners/sub-contractors based ontheir performance of service, regardless of whom and wherethey are.
� Customized Service & Specialized Demand: customization is
becoming more and more important in modern manufactur-ing, especially for SMEs. In a machine shop, specific cutter/machine tools are required for a particular job. With SCM, it iseasy to locate required facilities in the resource pool. There-fore, specialized objects are achieved without additionalinvestment on costly facilities and expertise.
� Facility Utilization: resource can be shared in a Cloud. Tech-
nical details and availability can be dynamically updated andpublished in the SCloud. Thus, manufacturing resources/cap-abilities can be better utilized. Production tasks can be easilybalanced between high-usage facilities and the low-usageones. From the user’s perspective, CUsers are able to choosethe available qualified providers for urgent jobs, or to wait forthe preferred facility in the queue. Therefore, the facilityutilization is improved by widely shared environment andreasonable schedule.
� Global Optimization: since services are broadcasted in the
Cloud, service solution can be improved and optimized basedon the virtualized service modules implemented in the Cloud.SCM predicts the service performance features beforehand,e.g., cost/time caused by preparing, machining, transportingand packing stages. So much so, the global solution is opti-mized based on particular factors or user’s preferences.
� Cost-Saving: by adopting the CManufacturing concept, the
manufacturing cost can be reduced. With the shared

MCapabilities available in the Cloud, optimized business solu-tion is easily found according to optimized results. Since thefeatures of SProviders are virtualized in the SC, it is more likelyto find supplier with better performance, cheaper labour,higher productivity, and better geographical location. As aconsequence of time-critical or cost-critical optimization stra-tegies, the performance of the service solution is predicted andimproved from a higher level and in a bigger scope of Cloud.Besides the cost of the service itself, the cost of strategicdecision is reduced as well. With the technical specificationshighly integrated in the SCloud, the cost of management,analysis, and comparison decreases, too.
� Better Enterprise Performance: when it comes to cost/time
management, ICMS improves not only the experience of theCUsers but also the enterprise’s performance as a CMServiceprovider. The MCapabilities are accessible in the Cloud, bring-ing more business opportunities. With the help of SCM, aSProvider is able to increase its production volume and reactrapidly to market changes.

4. Case studies

To evaluate the concept of ICMS, two case studies have beencarried out. MCapabilities are virtualized in the CManufacturingenvironment. The first case study shows the ability of integratingCMServices, and the second demonstrates how a CManufacturingenvironment can provide user with multiple options and improveenterprise performance at the same time.

Fig. 9. ICMS interface.


4.1. Cloud service segmentation

In the first example, a customer launches a customer interfaceremotely. The interface is developed in HTML format via theAdobe Dreamweaver package. It can be viewed within webbrowsers, e.g., IE, Safari and Firefox (Fig. 9). CUsers can start toselect CManufacturing Services from predefined categories andspecify service details. After logging on to the system, Usersare able to view their service history and the status of his orher services in-process. In this case, the user requires a seriesof services including product design, simulation, CNC milling,precision welding and grinding.

The service query descriptions and specifications are input viathe GUI. CIA collects these data and transmits the data into a XMLformat to enable Internet communications (Fig. 10). Note that theXML document which is transmitted from the client side iscompliant with the EXPRESS-G schemas but in a flat datastructure. This means that the meta-model only keeps theinformation of entity instance and temporarily suspends attri-butes and inheritance logics. The flat meta-model is utilized toenhance the data portability. CUsers are able to view and processthese data via general software tools, e.g., web browser, MicrosoftExcel, Access and so on. When the data comes back to the MCloud,it is mapped back to the structured model tree. Then, SCMcontinues to process the service process and searches for all thepossible solutions.

As a result, the SProviders which are able to meet the user’srequirements are summarized in Fig. 11. The service is con-structed by five service phases that are mapped to the five stagesof production. Multiple service providers are matched and stated

by the SCM. Due to the sequence of different routes, the serviceprocedure is presented in the flowchart. The service is illustratedas nodes and the transmitting methods as edges.

To optimize the solutions and provide a customized result,SCM analyses the feedback from the SCloud and evaluates theselections based on the user’s preference (e.g., cost-critical andtime-critical rules). The total cost includes the cost and theexpenses of freight between different providers. After SProvidersare found, all the possible transmitting services are detected asattachment services. The actual cost includes the cost of servicephase i (Cost(i)), and shipping cost that follows phase i (SC(i)) intotal.

cost¼ costðiÞþSCðiÞ ¼Xn

i ¼ 0

Cðiþ1Þ þXnþ1

i ¼ 0

SCi,ðiþ1Þ ð2Þ

In this case study, the SCM computes all of the possible pathsthrough service nodes and their edges. The sequential relation-ships between nodes can be represented as shown in Table 1.Value ‘‘1’’ indicates there is an operational sequence from node Xto node Y. The lowest cost combination is found and suggested tothe user. Since there are 36 combinations in this case, iterativealgorithm is deployed. For more complex business matters, neuralnetwork and genetic algorithm can used to improve the comput-ing efficiency.

Costopt¼Minimize½CostðPDj, Sk, CMl, PWm, GpÞ

þCostðSCjk, SCkl, SClm, SCmpÞ� ð3Þ

where PD is the product design; S is the simulation; PW is theprecision welding; G is the grinding.

Fig. 10. XML representation of a service request.

Phase1:ProductDesign

Phase2:Simulation

Phase3:CNCMilling

Phase4:PrecisionWelding

Phase5:Grinding

Shipping(Optional)

Shipping(Optional)

Shipping(Optional)

Shipping(Optional)

Shipping(Optional)Service Input

Start

PD1

PD2

PD3

S1

S2

CM1

CM2

CM3

PW1

PW2

G

F

FF

FFF

F

FF

F F

F

F

F

F

F

FF

F

F

F

F

F

F End

Service Output

PDS

CMPWGF

- Product Design- Simulation- CNC Milling- Precision Welding- Grinding- Freight

Fig. 11. Cloud service results pool.

Table 1Sequence matrix of cloud service phases.

From node X

To node Y Start PD1 PD2 PD3 S1 S2 CM1 CM2 CM3 PW1 PW2 G End

Start

PD1 1

PD2 1

PD3 1

S1 1 1 1

S2 1 1 1

CM1 1 1

CM2 1 1

CM3 1 1

PW1 1 1 1

PW2 1 1 1

G 1 1

End 1


Ta

ble

2C

lou

dse

rvic

em

ap

pin

gre

sult

sfr

om

SC

M.

Su

gg

es-

tio

n

(Co

st)

Su

gg

es-

tio

n

(Tim

e)

Se

rvic

e(a

ctiv

ity

)P

rov

ide

rS

erv

ice

seri

al

no

.

Act

ivit

yID

Pri

ce/e

ach

Qu

an

-tit

yT

ime

(we

ek

)C

ap

a-b

ilit

y/q

ua

lity

Pri

ce

(to

tal)

Tim

e

(to

tal)

Loca

tio

n

�P

rod

uct

de

sig

nC

AM

EX

S0

02

C0

01

11

00

12

nnn

11

00

2N

ort

hS

ho

re,

NZ

Pro

du

ctd

esi

gn

Ea

gle

Bu

rgm

an

nS

00

1E

00

69

00

13

nnn

90

03

No

rth

Sh

ore

,N

Z

�P

rod

uct

de

sig

nU

oA

Wo

rksh

op

S0

03

U0

01

80

01

2nnn

80

02

Au

ckla

nd

,N

Z

Fre

igh

tN

ULL

00

Da

taT

ran

s

Sim

ula

tio

nP

TC

Cre

oS

00

9P

00

11

00

01

2nnn

10

00

2P

TC

,U

S

��

Sim

ula

tio

nU

oA

Wo

rksh

op

S0

08

U0

03

70

01

1nn

70

01

Au

ckla

nd

,N

Z

Fre

igh

tN

ULL

00

Da

taT

ran

s

�C

NC

Mil

lin

gC

AM

EX

S0

19

C0

08

20

01

00

1nnn

20

00

01

No

rth

Sh

ore

,N

Z

�C

NC

Mil

lin

gB

EN

ZS

01

7E

00

41

60

10

01

nn

16

00

01

No

rth

Sh

ore

,N

Z

CN

CM

illi

ng

Uo

AW

ork

sho

pS

01

8U

00

42

00

10

02

nn

20

00

02

Au

ckla

nd

,N

Z

Fre

igh

tN

ULL

00

CA

ME

X-4

CA

ME

X

�Fr

eig

ht

NZ

PO

ST

S1

01

N1

01

50

01

0.5

nnn

50

00

.5C

AM

EX

-4B

EN

Z

Fre

igh

tN

ZP

OS

TS

10

2N

10

25

00

10

.5nnn

50

00

.5B

EN

Z-4

CA

ME

X

�Fr

eig

ht

NU

LL0

0B

EN

Z-4

BE

NZ

Fre

igh

tN

ZP

OS

TS

10

3N

10

36

00

10

.5nnn

60

00

.5U

oA

-4C

AM

EX

Fre

igh

tN

ZP

OS

TS

10

4N

10

46

00

10

.5nnn

60

00

.5U

oA

-4B

EN

Z

Pre

cisi

on

We

ldin

gC

AM

EX

S0

12

C0

04

60

10

02

nn

60

00

2N

ort

hS

ho

re,

NZ

��

Pre

cisi

on

We

ldin

gB

EN

ZS

01

3E

00

34

01

00

1nnn

40

00

1N

ort

hS

ho

re,

NZ

Fre

igh

tN

ZP

OS

TS

10

1N

10

15

00

10

.5nnn

50

00

.5C

AM

EX

-4B

EN

Z

Fre

igh

tN

ULL

0B

EN

Z-4

BE

NZ

��

Gri

nd

ing

BE

NZ

S0

10

E0

01

70

10

02

nn

70

00

2N

ort

hS

ho

re,

NZ


Computation under the time-critical rule is similar. Besides theperiod of service, results of transitions between SProviders cannotbe ignored. SCM traverses all the possible paths and finds themost sufficient solution in terms of time.

Time¼Xn

i ¼ 0

ðTðiþ1Þ þSTiðiþ1ÞÞ ð4Þ

where T is the time of service; ST is the time of shipping.

Timeopt¼Minimize½TðPDj, Sk, CMl, PWm, GpÞ

þTðSCjk, SCkl, SClm, SCmpÞ� ð5Þ

Eventually, the entire solution pool is detected and provided tothe user. The results are mapped into a general chart visible to theuser via the User Interface (Table 2). The optimized solution ismarked for the user. At this stage, the user is able to follow theSCM recommendation or customize the selection.

To recap, manufacturing services, from design to production,are virtualized and integrated within public CManufacturingarchitecture, without changing the existing organizational struc-tures of a manufacturing enterprise or company. The CMServiceparticipants are able to improve their enterprise performanceswhile maintaining their autonomy and competiveness. The parti-cipants are part of the integrated environment via contributingtheir valuable resources and services for mutual benefits.

4.2. Manufacturing service identification

When manufacturing enterprises are brought into the MCloud,the first step is to understand the MCapability of the enterprisewhich is then offered as a CMService. The second case studyconcerns a manufacturing service provider named BE, who hasfactories around the world. An example product is selected todemonstrate how industrial service is virtualized and implemen-ted in the MCloud.

The example product, ID T14859 Driver, is a ring-shape partwith multiple holes at different angles. An order of thirty parts areprocessed at BE Australia branch. BE has sites in Australia (BEAU)and New Zealand (BENZ), both of which are SProviders andcapable of machining this part. The MCapabilities of the two sitesare identified (Fig. 12).

Historically, the product data and process plans are documen-ted in different data semantics. Due to the geographical andcurrency differences, it is hard to evaluate machining tasks in thetwo sites. With the help of ICMS, the technical details are mappedfrom the flat chart to the modularized Cloud_Service_Specifica-tion model aforementioned. Compared with BEAU, there arestronger Hard Resources in BENZ, i.e., 9-axis multi-tasking CNCmachining centre, Sandvik T-Maxs U Trepanner cutters andSandvik saw blade cutters. The parameters and features aredescribed by the Entity Hard_Resource.

With the help of the 9-axis machining centre and itsbigger cutter magazine at BENZ, the clamping and setup time isdrastically decreased. Additionally, the service input is 316stainless steel round bar of diameter 300 mm. Originally, theremoving materials are in forms of chips that can only berecycled. Thanks to the Trepanner cutter, the result of themachining process is a hollow product with a cylindrical bar(diameter 130 mm) that can be utilized directly at BENZ. Thus,the actual material cost is drastically reduced. The cost marginsare mapped to the subtypes of Service_Cost. Based on thedata in this section, the actual service cost is summarized andprovided. Subsequently, BENZ is chosen to carry out the produc-tion task.

In this case study, ICMS serves as a community Cloud forthe organization. BE plays a role of service provider and user at

Fig. 12. Integrated CMService data.


the same time. Cloud-based virtualization mechanism providesexplicit integrating presentations of its service/products. It pro-vides clear knowledge of the operational processes. From theprovider’s perspective, BE is able to understand its services andreview/optimize them as required. As a CUser, the enterprise isable to choose the suitable service provider, which can be apartner/contractor connected by the MCloud or the enterpriseitself.

5. Conclusions

It is largely impossible to carry out some manufacturing taskswithout the support of suppliers, contractors and business part-ners. There is a need to connect manufacturers together to sharerisks, benefits, competiveness and costly resources. More thanjust deploying manufacturing-related software applications in theComputing Cloud, CManufacturing is an integrated solution thatprovides a pool of MCapabilities provided by Cloud participants.Manufacturers are integrated by either long-term arrangementsor short-term business objectives. In this paper, a Cloud-basedmanufacturing system, called ICMS is proposed to achieve acollaborative manufacturing environment. With the help of ICTtechnologies and MCapability identifications, it is possible torealize remote collaboration, coordination, and interaction amongparticipants.

To implement the CManufacturing concept, the key is toidentify existing manufacturing abilities and resources, to virtua-lize and implement them in the Cloud as trust-worthy manufac-turing services. A three-layer CManufacturing structure isproposed with a virtualization methodology describing

CMService, SProvider and CMService queries. Supervisionmechanism is developed for remote interaction between MCloudand Cloud clients that are categorized as either CU or EU. The casestudies evaluate the ICMS system. Operational processes areintegrated as CMServices from design to manufacturing. Optimi-zation and management can be further assisted by SCM in thefuture. Meanwhile, additional optimization tools can be possiblyadopted in the Cloud to improve the global performance of ICMS.

Efforts are made world-widely to realize faster time-to-marketand reduce the cycle time of a product. ICMS connects distributedcompanies by provisioning their MCapabilities as CMServices.With shared capacities, knowledge and competencies, it is possi-ble to for a business to take on more substantial projects that areotherwise not possible for one business to do. In a long term,more advanced manufacturing technologies can be brought intothe CManufacturing context to improve efficiency, interoperabil-ity, sustainability and flexibility of a manufacturing business.

References

[1] P Mell, Grance T. The NIST definition of cloud computing (draft). NIST specialublication, 800. NIST; 2011.

[2] Mell P, Grance T. Perspectives on cloud computing and standards. NationalInstitute of Standards and Technology (NIST). Information TechnologyLaboratory.2009.

[3] Apple, iCloud. (2012). Available: /https://www.icloud.com/S.[4] Amazon, Amazon Elastic Compute Cloud (EC2); (2012). Available: /http://

aws.amazon.com/ec2/S.[5] Google, Google App Engine—Google Code; (2012). Available: /http://code.

google.com/intl/en/appengine/S.[6] Microsoft, Windows Azure Platform _ Microsoft Cloud Services; (2012).

Available: /http://www.microsoft.com/windowsazure/S.

https://www.icloud.com/

http://aws.amazon.com/ec2/

http://aws.amazon.com/ec2/

http://code.google.com/intl/en/appengine/

http://code.google.com/intl/en/appengine/

http://www.microsoft.com/windowsazure/


[7] Oracle, Sun Cloud Developer Homepage; (2012). Available: /http://developers.sun.com/cloud/S.

[8] Xu X. From cloud computing to cloud manufacturing. Robotics andComputer-Integrated Manufacturing 2012;28:75–86.

[9] Tao F, Zhang L, Venkatesh V, Luo Y, Cheng Y. Cloud manufacturing: acomputing and service-oriented manufacturing model. Proceedings of theInstitution of Mechanical Engineers, Part B: Journal of Engineering Manufac-ture 2011;225:1969–76.

[10] Li B, Zhang L, Wang S, Tao F, Cao J, Jiang X, et al. Cloud manufacturing: a newservice-oriented networked manufacturing model. Computer IntegratedManufacturing Systems 2010;16:1–7.

[11] Terkaj, W, Pedrielli, G, Sacco, M.. Virtual factory data model, In: Proceedingsof the second international workshop on searching and integrating new webdata sources (VLDS 2012), Istanbul, Turkey; 2012.

[12] Meier M, Seidelmann J, Mezgar I. ManuCloud: the next-generation manu-facturing as a service environment. ERCIM News 2010;83:33–4.

[13] Ruiz OE, Arroyave S, Cardona J. EGCL: an extended G-Code language withflow control, functions and mnemonic variables. World Academy of Science,Engineering and Technology 2012;67:455–62.

[14] Yip ALK, Jagadeesan AP, Corney JR, Qin Y, Rauschecker U, Fraunhofer IL,, et al..System to support cloud-based manfuacturing of cutstomized products,In: Proceedings of the nineth international conference on manufacturingresearch ICMR; 2011.

[15] Wu L, Yang C. A solution of manufacturing resources sharing in Cloudcomputing environment, cooperative design, visualization, and engineering:.Lecture Notes in Computer Science 2010;6240:247–52.

[16] Hu CS, Xu CD, Cao XB, Fu JC. Study of classification and modeling of virtualresources in Cloud manufacturing. Applied Mechanics and Materials2012;121–126:2274–80.

[17] Luo YL, Zhang L, He DJ, Ren L, Tao F. Study on multi-view model for Cloudmanufacturing. Advanced Materials Research 2011;201:685–8.

[18] Luo YL, Zhang L, Zhang KP, Tao F. Research on the knowledge-based multi-dimensional information model of manufacturing capability in CMfg.Advanced Materials Research 2012;472:2592–5.

[19] Zhang, L, Guo, H, Tao, F, Luo, Y, Si, N. Flexible management of resource servicecomposition in Cloud manufacturing, In: Proceedings of the 2010 IEEE IEEM,IEEE; 2010. p. 2278–2282.

[20] Liu Q, Gao L, Lou P. Resource management based on multi-agent technologyfor cloud manufacturing, In: International conference on electronics, com-munications and control (ICECC), IEEE, Zhejiang, China; 2011. p. 2821–2824.

[21] Fan WH, Xiao TY. Integrated architecture of cloud manufacturing based onfederation mode. Computer Integrated Manufacturing Systems 2011;17:469–76.

[22] Laili Y, Tao F, Zhang L, Sarker BR. A study of optimal allocation of computingresources in cloud manufacturing systems. The International Journal ofAdvanced Manufacturing Technology 2012;63:1–20.

[23] Wang XV, Xu X. ICMS: a Cloud-based manufacturing system. In: Li W,Mehnen J, editors. Cloud manufacturing: distributed computing technologiesfor global and sustainable manufacturing. Springer; 2012 . [23]. in press.

[24] Buonanno G, Faverio P, Pigni F, Ravarini A, Sciuto D, Tagliavini M. Factorsaffecting ERP system adoption: a comparative analysis between SMEs andlarge companies. Journal of Enterprise Information Management2005;18:384–426.

[25] Gattiker TF, Goodhue DL. What happens after ERP implementation: under-standing the impact of interdependence and differentiation on plant-leveloutcomes. MIS Quarterly 2005;29:559–85.

[26] Ko DG, Kirsch LJ, King WR. Antecedents of knowledge transfer fromconsultants to clients in enterprise system implementations. MIS Quarterly2005;29:59–85.

[27] Liang H, Saraf N, Hu Q, Xue Y. Assimilation of enterprise systems: the effect ofinstitutional pressures and the mediating role of top management. MISQuarterly 2007;31:59–87.

[28] Wei CC, Chien CF, Wang MJJ. An AHP-based approach to ERP systemselection. International Journal of Production Economics 2005;96:47–62.

[29] Paulraj A, Lado AA, Chen IJ. Inter-organizational communication as a rela-tional competency: antecedents and performance outcomes in collaborativebuyer-supplier relationships. Journal of Operations Management 2008;26:45–64.

[30] Modi SB, Mabert VA. Supplier development: improving supplier performancethrough knowledge transfer. Journal of Operations Management 2007;25:42–64.

[31] Papazoglou MP, Van Den Heuvel WJ. Service oriented architectures:approaches. Technologies and Research Issues, the VLDB Journal 2007;16:389–415.

[32] Cherbakov L, Galambos G, Harishankar R, Kalyana S, Rackham G. Impact ofservice orientation at the business level. IBM Systems Journal 2005;44:653–68.

[33] Schmidt MT, Hutchison B, Lambros P, Phippen R. The enterprise service bus:making service-oriented architecture real. IBM Systems Journal 2005;44:781–97.

[34] Rauyruen P, Miller KE. Relationship quality as a predictor of B2B customerloyalty. Journal of Business Research 2007;60:21–31.

[35] Bernstein, PA, Melnik, S. Model management 2.0: manipulating richermappings, In: SIGMOD international conference on managment of data,ACM, Beijing, China; 2007. p. 1–12.

[36] ISO 10303–1: Industrial automation systems and integration – product datarepresentation and exchange – Part 1: Overview and fundamental principles;1994.

[37] ISO 14649-1. Industrial automation systems and integration – physicaldevice control – data model for computerized numerical controllers – Part1: Overview and fundamental Principles; 2003.

[38] Gielingh W. An assessment of the current state of product data technologies.CAD Computer Aided Design 2008;40:750–9.

[39] Zhang L, Luo YL, Tao F, Ren L, Guo H. Key technologies for the construction ofmanufacturing Cloud. Computer Integrated Manufacturing Systems 2010;16:2510–20.

[40] Allen RD, Harding JA, Newman ST. The application of STEP-NC using agent-based process planning. International Journal of Production Research2005;43:655–70.

[41] Monostori L, Vancza J, Kumara SRT. Agent-based systems for manufacturing.Annals of the CIRP 2006;55:697–720.

[42] Nassehi A, Newman ST, Allen RD. The application of multi-agent systems forSTEP-NC computer aided process planning of prismatic components. Inter-national Journal of Machine Tools and Manufacture 2006;46:559–74.

[43] ISO 10303-11 Industrial automation systems and integration – product datarepresentation and exchange – Part 11: Description methods: the expresslanguage reference manual; 2004.

[44] ISO 10303-21. Industrial automation systems and integration – product datarepresentation and exchange – Part 21: Implementation methods: clear textencoding of the exchange structure; 2002.

http://developers.sun.com/cloud/

http://developers.sun.com/cloud/

an interoperable solution for cloud manufacturing

Documents