Proceedings of the ASME 2013 International Manufacturing Science and Engineering Conference

MSEC2013
June 10-14, 2013, Madison, Wisconsin, USA

MSEC2013-1106

CLOUD MANUFACTURING: DRIVERS, CURRENT STATUS, AND FUTURE TRENDS

Dazhong Wu, Matthew J. Greer, David W. Rosen, Dirk Schaefer

The G.W. Woodruff School of Mechanical Engineering

Georgia Institute of Technology
Atlanta, Georgia, 30332, USA

ABSTRACT hold is dwarfed by that held by the global mass of people

Cloud Manufacturing (CM) refers to a customer-centric connected through globalization.
manufacturing model that exploits on-demand access to a Many engineering paradigms have evolved as the result of
shared collection of diversified and distributed manufacturing Globalization 3.0, some of which are mentioned by Tapscott
resources to form temporary, reconfigurable production lines and Williams (e.g., mass collaboration, self-organization). Of
which enhance efficiency, reduce product lifecycle costs, and the many paradigm shifts still in their infancy, Cloud
allow for optimal resource loading in response to variable- Manufacturing (CM) will be the focus of this paper. CM, as
demand customer generated tasking. Our objective is to present will be defined shortly, benefits from the share-to-gain
the drivers, current status of research and development, and philosophy as a wide number of manufacturing resources and
future trends of CM. We also discuss the potential short term expertise can be combined to provide consumers with enhanced
and long term impacts of CM on various sectors. experiences. Many researchers have offered different
definitions of CM, including Tao et al. [23], Xu [4], Wu et al.
1 INTRODUCTION [5], and Schaefer et al. [6]. Most notably, Xu discerns between
The force of globalization has served to instantaneously two forms of cloud manufacturing: the introduction of cloud-
connect people from all across the globe, bringing with it game- computing technologies into the manufacturing environment
changing opportunities to share knowledge and expertise to and cloud manufacturing. The latter is a replication of the
benefit in a collective manner (so called share-to-gain). cloud-computing environment using physical manufacturing
Friedman [1] explains that the latest globalization phase, which resources in lieu of computing resources – this idea will be the
he coins Globalization 3.0, began around the year 2000 and was focus of this paper. Using the work of NIST and Smith as a
enabled by the expansion of the internet on a global basis foundation, the following working definition of Cloud
during the dot-com boom. According to Friedman, Manufacturing is offered:
Globalization 3.0 is defined by individuals and small groups Cloud Manufacturing (CM) is a customer-centric
from across the globe collaborating in areas once dominated by manufacturing model that exploits on-demand access to a
less-connected western economies. shared collection of diversified and distributed manufacturing
Tapscott and Williams [2] explain that the advent of the resources to form temporary, reconfigurable production lines
internet has led to the development of cooperative collaboration which enhance efficiency, reduce product lifecycle costs, and
networks that have resulted in a power-shift from the once allow for optimal resource loading in response to variable-
mighty hierarchical business models. These traditional business demand customer generated tasking.
models, according to the authors, can no longer sustain The key drivers behind CM and associated metrics are
successful innovation: “In an age where mass collaboration can listed in Table 1. First, one of the main reasons for the adoption
reshape an industry overnight, the old hierarchical ways of of CM by manufacturing enterprises is the emerging
organizing work and innovation do not afford the level of outsourcing and crowdsourcing models in design and
agility, creativity, and connectivity that companies require to manufacturing. These models help small and medium
remain competitive in today’s environment.” Simply put, enterprises (SMEs) to reduce cost by outsourcing their
industry is going to have to rethink the traditional models of operations and support services. Second, one of the
business operation, as the amount of internal expertise they distinguishing characteristics of CM is agility allowing for
reconfiguring the products and associated manufacturing

1 Copyright © 2013 by ASME

systems to accommodate the rapid changing customer needs. environments, CM systems must possess flexibility
Third, CM allows enterprises to quickly scale up and down, and agility which will help reduce manufacturing cost
where manufacturing cells, general purpose machine tools, and time to market.
machine components (e.g., standardized parts and assembly),  For CM to be embraced by service consumers and
material handling units, as well as personnel (e.g., designers, providers, both of them must have an innovative
managers, and manufacturers) can be added, removed, and business model so that the unique value can be added
modified as needed to respond quickly to changing by CM that traditional manufacturing paradigms
requirements. It helps to better handle transient demand and cannot offer.
dynamic capacity planning under emergency situations incurred  As more and more service consumers and providers
by unpredictable customer needs and reliability issues. Fourth, get involved in CM, it is crucial for the cloud to have
CM enables manufacturing resources to be pooled to serve CM the searching capability for collaborator selection
consumers in a pay-per-use fashion which helps resource which helps the consumers find the suitable suppliers
sharing between service providers. Fifth, there is an increasing in the CM network.
amount of data and unstructured information about product  Due to the complexity of the CM systems, the
design and manufacturing collected by information and implementation architectures are needed to facilitate
communication technology (ICT) systems. CM has the the design and development of CM systems.
potential to allow the collected information to be processed, The work that follows in no way defines the full extent of
managed, and shared within and across organizations in the CM any particular field of study; rather, it documents those aspects
networks. Last but not the least, more companies have most important to the enablement of CM.
embraced social networks as a means of interaction and
communication with users and customers [26]. Research and 2.1 Automation
development in social product development can help us grasp Stouffer et al. [10] discuss typical industrial control
fundamental insights into collaboration relationships in the systems (ICSs) utilized in both process based and discrete-
socio-technical network formed in CM systems. based manufacturing environments, including Supervisory
Table 1 The drivers and metrics for CM Control and Data Acquisition (SCADA), Distributed Control
Systems (DCSs), and Programmable Logic Controllers (PLCs).
Drivers Metrics SCADA is a type of control system used to control
manufacturing enterprises which are distributed over a large
Economy Cost
area, and is typically utilized in the gas and utilities industries.
Agility Response time to market changes DCSs are used to control industrial process variables around a
Scalability Throughput, utilization set target, and are common in process intensive industries.
Resource sharing Amount of pooled resources Finally, PLCs are computer based logic devices that control
Information sharing Semantic similarity, interoperability equipment and processes, and are often employed as part of a
Social sphere Centrality and network density DCS system.
Programmable Automation Controllers (PACs) as
2 CURRENT STATUS described in [15] are a relatively new form of ICS which focus
As more attention paid to CM, the current state of the field on emerging issues that limit ICS such as network connectivity,
must be collected from many different specialties which, in device interoperability, and enterprise data integration.
their combination, provide a foundation for the advancement of According to [15], PACs feature modular designs, open (non-
CM. Several fields of study were used to compile the following proprietary) architectures, and the ability to monitor multiple
current status information, including but not limited to signal types such as analog, digital, and serial.
automation, service composition, flexibility and agility, One of the most advanced control systems demonstrated in
business model, collaborator selection, and implementation industry is the Siemens Totally Integrated Automation (TIA)
architectures. system, which offers a wide range of control technologies in
 Automation technologies will facilitate inter- and both SCADA and DCS environments [16]. The TIA system is
intra-factory communication and collaboration in CM based upon an open system architecture, which promotes
environments, allowing for automatically executing modularity and interoperability with existing assets.
manufacturing tasks generated by CM consumers. Much work has recently been done to establish open
 As service consumers focus on their core business and standards promoting technology connectivity and machine-to-
outsource other application series over Internet, machine communications. Developments in open architecture
service composition becomes critical which deals with standards and communication protocols will serve to facilitate
selecting and integrating inter-organizational and automation through promotion of “plug and play” technologies
heterogeneous services in CM environments. which can be offered from a wide variety of sources. The
 As the manufacturing environment is very dynamic MTConnect Institute [21] has developed open and royalty free
and ever-changing, in order to survive in such communication standards based upon the Extensible Markup

2 Copyright © 2013 by ASME

Language (XML). These standards allow for machine-to- Tao et al. [23] present that the cloud manufacturing
machine communications and promote interoperability. environment is in-part enabled by the creation of
Research regarding machine-to-machine communication is Manufacturing Cloud Services (MCSs). MCSs are cloud
also common in the academic realm, as shown by references services that are formed when manufacturing resources are
[20] and [9]. CyberOPC is a dedicated protocol developed for virtualized and encapsulated. They explain that MCSs can be
communication with CNC machines over public networks [20]. categorized and combined into related manufacturing clouds,
The use of STEP-NC is discussed as a communication language from which consumers can select particular MCSs to form their
between the shop floor and the plant scheduling level in [9]. required production facility.
As a result, to advance CM, the challenges associated with Therefore, the challenge in terms of service composition is
automation are (1) developing the unified standards for how to automate service composition through formal
describing the function, structure, behavior of the manufacturing resource representations. Some ontology tools
interconnected equipment in the cloud which will enable the can perform automated reasoning using the ontologies, and thus
devices to have self-description, self-configuration, and enable intelligent discovery services such as semantic search,
context-awareness capabilities; (2) designing wireless smart information retrieval, and service matching.
sensor network based monitoring systems to keep track of real-
time information from the manufacturing environment for 2.3 Flexibility and Agility
scheduling and optimized resource allocation. Hao et al. [14] state that advanced manufacturing systems
are geared towards agility, that is, they are adaptive to changing
2.2 Service Composition market conditions and variable customer requirements. Panchal
Service composition deals with the ability to provide useful and Schaefer [17] define agility as the ability to successfully
manufacturing services to consumers based upon the available and quickly adapt to changes in the operating environment,
manufacturing resources held by the resource providers. That both expected and unexpected. They further state that agility in
is, service composition deals with the ability to extract useful the manufacturing realm often deals with the ability to quickly
services from available resources and combine them when adapt a manufacturing resource to produce a different
needed for value-added operations. component or assembly. Implementation of agility and
Xu [4] discusses the creation of cloud manufacturing flexibility in cloud manufacturing systems differs significantly
services, which are virtualized manufacturing resources made across the literature.
available to consumers through the cloud. The formation of
Table 2 Flexibility categories
cloud manufacturing services is enabled by the ability to
identify, virtualize, and package both tangible and intangible
resources. Xu presents numerous methodologies of identifying Flexibility Type Implication
distributed resources, including such technologies as RFID, Task Flexibility RSCs can be constructed to adapt to
wireless sensor networks, and Global Positioning System many different tasks
(GPS), among others. The method of resource virtualization, Flow Flexibility Many RSC paths can be used to reach
according to [4], depends upon the form of resource being the required final condition
virtualized. Computational and knowledge resources would be Resource Service Single resource services can complete
virtualized in a similar manner to that used in cloud computing Flexibility many different tasks
technologies, and hardware resources would be converted into Quality of Service RSC can maintain a certain QoS, which
virtual machines using agent based technologies for distributed (QoS) Flexibility is flexible
control and communication. Packaging resources and making Correlation RSC can adapt to changes in correlations
them available as cloud based services, according to Xu, would Flexibility among resources
be accomplished through description languages. One example
of a cloud manufacturing service provided by Xu is STEP Zhang et al. [3] discuss how the lifecycle of RSCs can be
Resource Locator (STRL), which uses an URL, Action, and affected by numerous factors. These authors argue that based
Query to identify a machine and task it with some requested upon the possible RSC interruptions, five forms of RSC
service instructions. flexibility are required for maximum system adaptability: task,
Zhang et al. [3] define Resource Service Composition flow, resource service, QoS, and correlation. These flexibility
(RSC) as the integration of existing resources to form categories are summarized in Table 2. The management of RSC
composite services which can be used to address complex is promoted through the adoption of a Flexibility Management
manufacturing tasks. According to them, the RSC has a four Architecture, which is composed of three functional modules.
stage lifecycle (design, deployment, execution, and post- The function module constructs the RSC, optimizes it, and
processing) which can be affected by numerous variables. begins the execution phase. During RSC execution, the
The RSC lifecycle is initiated and maintained through a tri- Monitoring module monitors those variables which affect the
modular system which executes the RSC, monitors for factors RSC lifecycle, and transmits information regarding abnormal
affecting lifecycle, and adjusts the RSC based upon system changes to the coordination module. The coordination module
changes.

3 Copyright © 2013 by ASME

then invokes corresponding adjustments to the RSC to ensure develop an improved plating process. This process, created
continued operations. collaboratively, would be the equivalent to foreground rights
Hence, the challenge we are faced with in terms of resulting from a joint venture. Through proper negotiations
flexibility and agility is developing an effective resource between CM parties, both background IP interests can be
virtualization framework for CM which can facilitate protected and agreements can be made as to the use of
manufacturing resource sharing, reduce manufacturing foreground rights.
reconfiguration cost, and accelerate product time to market. Wagner et al. [18] offer further research regarding value
management in collaborative environments. Consider their
2.4 Business Model definition of value creation and appropriation:
In a broad sense, CM business models will need to support “From a conceptual perspective, value creation and value
collaboration and cooperation to an unprecedented extent, as appropriation represent two sides of the same coin. Value
the mere survival of CM value chains will be reliant upon creation entails the total net value (i.e., total outcomes minus
efficient and effective group action. Social psychology has total inputs) created in a collaborative effort among exchange
offered numerous theories surrounding cooperation and partners. Value appropriation depicts the net value that a focal
collective decision making, all of which will help develop firm claims successfully. Value creation is a win-win scenario;
effective business models. Two of the most relevant social value appropriation means that a larger value slice for one party
psychology theories for future CM environments are equity diminishes the remaining slice for the other partner.”
theory and game theory. Equity theory [7] deals with why Maximizing value creation while appropriating the reward will
individuals participate in groups and how they react when be the main focus of effective CM business organizations. They
outcomes are disproportionately distributed. Equity theory is offer a number of hypotheses regarding the link between value
composed of four propositions as shown in Figure 1. creation and appropriation, and how these concepts affect the
Equity theory is important to the development of CM attitudes and behaviors of participating parties. Their work
business models because it enforces the need for just and fair builds upon the equity theory, which explains how collaborative
reward sharing amongst CM collaborators. partners measure their inputs into a project against the benefits
Game theory [8] deals with how rational individuals make they receive, and also expresses how misappropriation of
decisions in mutually interdependent roles. Game theory can be reward can negatively impact business relationships. Using data
non-cooperative, in which individuals can act together but are from 186 manufacturing companies, they showed support for
not bound by formal agreements, or it can be cooperative, all but one of their eleven hypotheses. The three main
where formal cooperation agreements are utilized. Game theory conclusions from them are (1) satisfaction is most highly driven
will help understand the motivations in cooperative by value appropriation; (2) collaborators compare their awards
environments and will help develop business models which with those of others; (3) the open and frequent exchange of
further the interest of the group. information can ease tensions between competitors.
Moreover, the CM environment will utilize relationships Thus, one of the critical issues we need to address is
that resemble those of joint ventures or collaborations. Parker identifying effective operation modes (i.e., private, public, and
[24] explains that joint ventures involve the formation of a legal hybrid clouds) and product/service co-creation frameworks by
entity separate from the parties coordinating the venture, while applying game theory into CM which can create the win-win
collaborative relationships involve two or more parties working situation for both service consumers and providers in the cloud.
together under contractually enforced terms. The purpose of
both joint ventures and collaborations is to share information 2.5 Collaborator Selection
and expertise in order for all parties involved to do something One of the key issues in CM is partner selection which is to
they otherwise could not. Parker states that intellectual property find suitable partner candidates who can offer certain products
considerations occur throughout a four stage life cycle (pre- or services in time with high quality at low price [29]. Lomas
contractual, formation, duration, and termination) of [19] suggests that developing a complete set of competencies
collaborative relationships. Throughout the entire collaborative opens new markets and allows for adaptability to change. When
relationship, the most important issues about the use and searching for collaborative parties, Lomas suggests considering
control of background and foreground rights. Background competencies, previous collaborations, and industry experience.
rights, according to Parker, are those that each company holds Huang [25] points out that the formation of a distributed
prior to the relationship and intends to contribute to the venture. manufacturing network requires companies to perform three
Foreground rights are those generated through the action of steps. First, the company must identify possible collaborators
both parties throughout the length of the venture [24]. Similar and their interests (similar to the work of Lomas). Second, they
background and foreground rights will exist in the CM must know how to relate the collaborators’ interests with those
environment. For example, a CM network may include of their own firm. Finally, the company must select the best
numerous plating houses, each with existing plating process collaborators who can work with the company to fulfill
specifications (background rights) that they will want to protect common interests. Choy et al. [28, 30] present intelligent
as intellectual property. At the same time, the CM environment supplier management tool for selecting and benchmarking
may create the need for these plating houses to collaborate to suppliers under outsource manufacturing environment using the

4 Copyright © 2013 by ASME

case-based reasoning and neural network. Viswanadham and infrastructure, and the centralized interface enables the system
Gaonkar [27] develop a mixed-integer programming model for to function as a whole. Human assets include consumers,
integrated partner selection and scheduling in an Internet- producers, and managers. The communication assets proposed
enabled dynamic manufacturing network environment. Yoo et include a communication network (internet), network security,
al. [31] introduce an innovative cyberinfrastructure-based and two interfaces for communicating with the human and
framework which allows designers to identify parts suppliers in manufacturing process asset groups. The manufacturing process
global and virtual environments more precisely. asset group is composed of hardware and software resources
Hence, the challenge in terms of collaborator selection is used in the CBDM environment.
how to model the socio-technical network formed in CM and Tao et al. [23] propose a four stage cloud manufacturing
how to identify key actors and potential collaborators in the model where manufacturing resources are controlled through
CM network. the internet through intelligent monitoring systems. These
resources are then virtualized and encapsulated into
2.6 Implementation Architectures and Frameworks Manufacturing Cloud Services (MCSs). These MCSs, in
Architectures and frameworks for implementation of CM contrast to the actual physical resources they represent, can be
have been presented by numerous authors. These proposed accessed and invoked in the cloud. After creation of many
structures vary in their complexity, maturity, and level of different MCSs based upon the manufacturing resources
demonstrated potential, yet many have similar characteristics. available, the MCSs are categorized and organized into
Xu [4] proposes a four layer CM framework consisting of a manufacturing clouds. Users can then search the manufacturing
manufacturing resource layer, a virtual service layer, a global clouds for services and combine MCSs to fit their needs. In
service layer, and an application layer. According to Xu, the similar fashion to other architectures presented, that proposed
Manufacturing Resource Layer contains the physical in [23] consists of manufacturing resources and abilities at the
manufacturing resources and capabilities of the shop floor, lowest level. These resources are then virtualized and managed
which are ultimately provided to the customer in Software-as-a- in a cloud environment, and then made available to consumers
service (SaaS) and Infrastructure-as-a-service (IaaS) delivery through an application layer. The functional layers of the
models. The virtual service layer identifies, virtualizes, and architecture are facilitated by the layers of knowledge, cloud
packages the resources as cloud manufacturing services, which security, and a network such as the internet. They also state that
are then managed using the Global Service Layer (GSL). The cloud manufacturing platforms can be public, private,
GSL, depending upon the task demanded, can operate in both community, or hybrid environments.
partial and complete services modes. In the partial service Mahesh et al. [11] propose an agent based framework for
mode, the GSL does not handle all CM related activities – distributed collaborative manufacturing. This idea is based
rather, the resource providers take some control of the process upon software agent technology where agents are programs that
flow and the GSL helps administratively manage the CM can act logically to perform given tasks with minimal user
activities. The complete service mode, however, coordinates input. They propose a system composed of many different
and manages the entire CM activity. Most importantly, the GSL agents that can share common resources and distribute tasks as
is a cloud platform and provides services using the Platform-as- required for success. According to them, the structural
a-service (PaaS) model. Xu discusses the Application layer, framework provides defined relationships between agents,
which provides the user-resource exchange portal. Through the while communication ontology facilitates cooperation.
Application layer, the user can construct manufacturing Knowledge Query and Manipulation Language (KQML) is
applications from the virtualized manufacturing resources. proposed as a communication protocol amongst agents.
Wu et al. [5] propose a Cloud Based Design and Additionally, they suggest that a central management agent be
Manufacture (CBDM) model composed of a cloud consumer, used to coordinate the efforts of all other agents.
cloud provider, cloud broker, and cloud carriers. The cloud Alvares and Ferreira [12] also propose an agent
consumers serve the obvious role of utilizing the cloud’s environment. These authors propose a 3 level (design, process
services, while the providers have the equally obvious role of planning, and manufacturing) environment managed by
providing services in the cloud. The cloud broker is an distributed agents. The use of KQML is proposed as a means of
intermediate party between the consumers and providers, and inter-agent communications, and the use of a central
also manages the use, performance, and delivery of services. management agent is also proposed. Hao et al. [16] also
The cloud carriers enable the exchange of services between propose the use of agent technologies to manage manufacturing
providers and consumers through the provisioning of transport at the inter-enterprise level, the intra-enterprise level, and the
networks. shop floor level. The agent architecture proposed is facilitated
Schaefer et al. [6] propose a Distributed Infrastructure with by Web Services and Agent Communication Language (ACL).
Centralized Interfacing System (DICIS) as a CBDM structuring Therefore, the challenge in terms of implementation
architecture. The DICIS is composed of three asset groups architectures and frameworks is how to implement the concepts
(human, communication, and manufacturing process) bounded of SaaS, PaaS, IaaS, and Hardware-as-a-service (HaaS) in CM.
by a centralized interface and a distributed infrastructure. The
three asset groups are combined together in the distributed

5 Copyright © 2013 by ASME

 Current Supply Chain CBM Supply Chain

OEMs (Users)
Original Equipment
Manufacturers (OEMs)
Requirements
Suppliers
Requirements

Tier I Suppliers Requirements

Cloud Based Applications

Derived
Requirements Derived
Requirements
Sub-Tier Suppliers
Manufacturing Service &
Resource Providers

Figure 2: Comparison between current and CM supply chain

3 FUTURE TRENDS numerous production paths that could lead to the desired object.
Finally, the application layer is responsible for locating the
3.1 Cloud Manufacturing Environment required resources, pending them to the engineering job, and
The CM environment will consist of interaction between managing resources in the event of a service interruption.
three groups: (1) the users (consumers), (2) application
providers, and (3) physical resource providers. The needs of 3.1.3 Physical Resource Providers
users will be matched with the capabilities of resource Physical Resource Providers (PRPs) own and operate
providers through the application layer, which will serve to manufacturing equipment, including but not limited to
create unique manufacturing protocols, develop sequencing and machining technologies, finishing technologies, inspection
planning data, and locate and manage those resources held by technologies, packaging technologies, and testing resource.
the resource providers. These PRPs are not limited by geographic location; rather,
PRPs join the CM network based upon their expertise alone.
3.1.1 Users Ideally as a whole, the PRP network would represent every type
Users are the consumers in CM; these individuals or of manufacturing capability available in the marketplace,
groups have the need to manufacture something, but do not offering users instantaneous access to manufacturing
possess the capabilities to do so, or stand to gain a competitive capabilities provided as a service. The input to the PRP group is
advantage by utilizing CM despite their own capabilities. the manufacturing data created by the cloud based applications,
Users can range anywhere from hobbyists to large OEMs – any and the output is the finalized product.
group that can generate engineering requirements to be used in
a manufacturing setting can participate in CM partnerships. 3.2 Key Characteristics
These engineering requirements, which describe the desired
object and its final conditions, are provided to the cloud based 3.2.1 Customer Centricity
application layer for interpretation. 21st century Industry is dominated by hierarchical supply
chains in which requirement originating parties flow down
3.1.2 Cloud Based Applications product level requirements to suppliers, who can then engage
The cloud based application layer takes user requirements sub-tier suppliers to assist with the product development
and interprets them to develop data required for production of process. As illustrated in Figure 2, a classic example of this
the desired objects. For example, a user desired product may relationship is that of an Original Equipment Manufacturer
require the development of a CNC tool path program and (OEM) who develops product level requirements from the
process paperwork to achieve a final desired plating condition – perspective of technology function and integration. These
these would be created by the cloud based applications. requirements are then contractually enforced with a first-tier
Furthermore, production planning and sequencing can be supplier, who can then contract out sections of the work to sub-
carried out through automated applications that determine the tier suppliers based upon the nature of the work and core

6 Copyright © 2013 by ASME

competencies. While often these relationships can be fruitful inherent system flexibility will be utilized to ensure even load
for all parties involved, the opportunity to enhance the sharing across equivalent or interchangeable manufacturing
consumer experience (reduce costs, improve quality, etc.) are resources. For example, if manufacturing resource “A” is more
severely limited by their rigid nature. Furthermore, when heavily utilized than others in the network, yet the desired
traditional supplier relationships prove to be undesirable, they process can be performed by combining manufacturing
can often prove to be difficult and costly to dissolve. resources “B” and “C”, the CM environment will realize and
In the CM environment, manufacturing supply chain capitalize upon this alternative to avoid excessive loading of
relationships will be customer-centric, defined by enhanced manufacturing resource “A”. An example of such a scenario
efficiency, reduced cost, increased flexibility, and improved would be the requirement for a 5-axis CNC machine, when the
capabilities. These benefits will be derived from the creation of combination of vertical and horizontal mills could be used to
flexible manufacturing sequences enabled by the pooling of process the job.
resources from many different sources. Solutions will be
customer, or even task, specific, as the cloud based application 3.2.4 Shared Burden, Shared Benefit
layer can be used to generate numerous options for the users Traditional business organizations rely upon a tiered
based upon their specifications (the user would be allowed to structure of control, which acts together to create value.
specify key aspects of the desired job, such as cost, lead time, Business organizations often vary widely across industries, and
and quality, and different choices that fit within those ranges can even be different across corporations within the same
would be provided for consideration). The key goal of a CM industry. The organization of a business often defines a
environment is linking users, with needs, to resource providers company as much as does the product or service it offers. Mari
who can fulfill those needs while meeting cost, schedule, and Sako [13] explains that business models define business
quality objectives of the user. operation: “a business model articulates the customer value
proposition; it identifies a market segment; it defines the
3.2.2 Temporary, Reconfigurable, Dynamic structure of the value chain; it specifies the revenue generation
Another distinguishing characteristic of CM is the mechanisms; it describes the positioning within the value
dynamic, flexible nature of resource provisioning. CM network or ecosystem; and it also elaborates on competitive
production lines are meant to be temporary in nature, allowing strategy by which the firm gains and holds advantage over
for the production of small lots but not excluding the rivals.” A business model is the argument as to why the
opportunity for longer production runs as well. The ability to company will succeed – it explains critical things such as who
reconfigure and repurpose manufacturing resources allows for the customers are, why they care about your product or service,
high efficiency, minimized down time, and instant demand how you are going to add value to the product, and how you
response. will make money.
System flexibility will rely upon the ability to rapidly re- The organizing business models that will someday define
configure and repurpose manufacturing equipment across CM, while not unprecedented altogether, will require a shift
multiple dispersed manufacturing sites with minimized down from traditional business models of today to ones based on the
time. To accomplish such a task, a high level of automation will share-to-gain philosophy. Value chains, which describe how
be required to ensure that the division of tasks can be properly value is added to a product, will be highly flexible in CM.
flowed down to the shop floor with minimal effort. The Value will be added by resource providers sharing expertise and
integration of automation, which in many industries is already collaborating to provide users with the products they desire
present today, does not necessarily imply the absence of human while utilizing less resources through efficient processes. CM
beings. Depending upon the application, the entire will require the formation of new business models altogether
manufacturing process may be automated, and in other (by all vested parties), and will require propositions as to what
instances humans will still interact as a measure of quality value the customer will receive, what market there is for such a
assurance and error prevention. business and so on. The appropriate business model for CM
In order to flow manufacturing requirements from the may be difficult to determine when it comes to value chain
cloud to automated resources, ICSs will be required. These structure and revenue models. In traditional business models,
control systems will act as the central nervous system, the value chain and revenue models are firmly defined – each
monitoring and controlling resources at the shop floor level to value adder is separated from the others, and they are
ensure multi-resource cooperation. The ICSs will coordinate compensated based upon the value they can add to the product.
and distribute tasks amongst manufacturing locations, ensuring In a CM environment, collaboration between suppliers will be
compatibility of efforts and final products. required to successfully complete a project. How will value
added be determined when three different manufacturers
3.2.3 Demand Driven, Demand Intelligent combine resources to complete a build-to-print order? Will the
Like any manufacturing entity today, the extent to which overall value of the final parts be divided evenly between
the CM environment is exercised will be driven by user suppliers, or will it be shared based upon time and resources
demand. Unlike traditional manufacturing enterprises, however, spent? These are the questions that will determine how value
the CM environment will be “demand intelligent” in that the chains are structured and how wealth sharing will occur.

7 Copyright © 2013 by ASME

 Marketing & Service
Reduce time-to-market

Improved service quality Customer co-creation

Enhanced customer needs elicitation

Manufacturing

Improved resource sharing

SECTORS

Rapid prototyping Distributed manufacturing

Reduced cost
Engineering & Design

Ubiquitous access to design information

Improved efficiency Collaborative design

Affordable computing resources

Short term TIME Long term

Figure 3: The potential impact of CM across sectors

CM will likely cause a shift in the revenue models Engineering design:

currently used by design firms and manufacturers alike. The In the short term, the benefits of CM on engineering design
introduction of the cloud will cause a shift in how value is are ubiquitous access to design information, improved
added to the product, as the cloud will take over some of the efficiency, and affordable computing resources. In the long
activities that contribute to the revenue models of both the users term, the impact area is collaborative design which is to support
and resource providers. Firstly, the cloud will introduce a engineering design in geographically dispersed environments.
change in how users calculate the cost of doing business. In the past two decades, the most important research works
Secondly, the cloud will remove some opportunity for service in collaborative design are web-based design and agent-based
providers to add value to products, requiring them to adjust design. The architectures for web-based collaborative design
their business models accordingly. can be classified into three categories: thin server and thick
The implementation of CM business models may also client, thin client and thick server, and peer-to-peer which are
struggle due to data ownership rights. Traditionally, data rights enabled technologies such as the Web, HTML, ActiveX, and
are easily understood – design authorities own the rights to the CORBA. In agent-based design, the agent technology allows
product designs, and manufacturers of those designs own the developers to focus on objects rather than functions, providing
manufacturing data that is used to produce it. Consider now that applications that are modular, decentralized and changeable.
cloud based applications will be used to generate much of the However, both web and agent-based approaches are lack of
value once produced by manufacturers (tool path programming, socio-technical network, mass collaboration, and inter-
process planning, etc.). Users may claim that data produced by connected design knowledge pool capabilities which CM may
the cloud is their property because they paid for access to the have the potential to possess. An industry application example
cloud-based applications. Those firms managing the cloud- is Quirky [32]. The Quirky business model incorporates the
based applications will certainly argue that it is their property originating designers into the wealth sharing model and
for distribution to whomever they like. The physical resource provides them with a portion of the profits that their products
layer might also try and argue it is their data, because without yield.
their expertise it would be of little use.
Manufacturing:
In the short term, the benefits of CM on manufacturing are
4 POTENTIAL IMPACTS
improved resource sharing, rapid prototyping, and reduced cost.
As illustrated in Figure 3, we envision that the possible
In the long term, the impact area is distributed manufacturing.
impact of CM on three key sectors including engineering
Although current research and developments in distributed
design, manufacturing, as well as marketing and service.
manufacturing bring a significant contribution, the research

8 Copyright © 2013 by ASME

work achieved so far are not sufficient to satisfy the needs of REFERENCES
modern manufacturing enterprises. One of the critical issues is [1] Friedman, T. L. (2005). It's a flat world, after all. The
still not fully addressed yet is scalability. However, CM has the New York Times, 3, 33-37.
potential to offer rapid scalability in some situations at certain [2] Tapscott, D., & Williams, A. D. (2008). Wikinomics:
levels, such as manufacturing cells, general purpose machine How mass collaboration changes everything. Portfolio
tools, and standardized machine components. For example, 3D Trade.
printing technology does not require tooling, allowing the cloud [3] Zhang, L., Guo, H., Tao, F., Luo, Y. L., & Si, N. (2010).
service providers to rapidly scale up and down manufacturing Flexible management of resource service composition in
capacity. In addition, the 3D printers connected in the cloud cloud manufacturing. In Industrial Engineering and
also help rapid tooling which makes rapid scalability possible Engineering Management (IEEM), 2010 IEEE
for traditional manufacturing processes requiring tools. One of International Conference on (pp. 2278-2282).
the industry application examples is Ponoko which is an online [4] Xu, X. (2012). From cloud computing to cloud
service for distribute and on-demand manufacturing [33]. manufacturing. Robotics and computer-integrated
Ponoko developed a personal factory software platform for manufacturing, 28(1), 75-86.
users to produce small-scale products through 3D printing, [5] Wu, D., Thames, J.L., Rosen, D., & Schaefer, D. (2012).
CNC machining, laser cutting services. The Cubify.com 3D Towards a cloud-based design and manufacturing
online printing service is another good example, which allows paradigm: looking backward, looking forward.
cloud consumers to produce parts through any mobile device Proceedings of the ASME 2012 International Design
using their online 3D printing service without purchasing 3D Engineering Technical Conference & Computers and
printers. Information in Engineering Conference, Paper Number:
DETC2012-70780, Chicago, U.S.
Marketing and service:
[6] Schaefer, D., Thames, J.L., Wellman, R.D., Wu, D., Yim,
In the short term, the benefits of CM on marketing and
S., & Rosen, D. (2012). Distributed collaborative design
service are reduced time-to-market, improved service, and
and manufacture in the cloud - motivation, infrastructure,
enhanced user experience. In the long term, the impact area is
and education. 2012 American Society for Engineering
customer co-creation.
Education Annual Conference, Paper #AC2012-3017, San
In order for manufacturing enterprises to create value
Antonio, U.S.
through collaboration, there is an increasing need to establish a
[7] Hatfield, E., Walster, G. W., & Berscheid, E. (1978).
new form of information, knowledge and resource sharing
Equity: theory and research. Boston, MA: Allyn and
mechanism that emphasizes the generation and realization of
Bacon.
various product stakeholders’ value. CM has the potential to
[8] Romp, G. (2001). Game theory: introduction and
create new marketing channels for information and resource
applications. OUP Catalogue.
sharing which will transform the traditional product realization
[9] Li, P., Hu, T., & Zhang, C. (2011). A unified
process into a value co-creation process. Specifically, the co-
communication framework for intelligent integrated CNC
creation process enhanced by CM can engage customers,
on the shop floor. Procedia Engineering, 15, 840-847.
designers, manufacturing engineers, and production managers
[10] Stouffer, K., Falco, J., & Scarfone, K. (2008). Guide to
to communicate with each other through social media such as
industrial control systems (ICS) security. NIST Special
Facebook, Twitter, Blogs, and online forums. The best industry
Publication, 800, 82.
practice is Salesforce.com for its well-known customer
[11] Mahesh, M., Ong, S. K., Nee, A. Y. C., Fuh, J. Y. H., &
relationship management services such as sales cloud, service
Zhang, Y. F. (2007). Towards a generic distributed and
cloud, and collaboration cloud. For instance, the sales cloud
collaborative digital manufacturing. Robotics and
allows its users to access sales data anywhere through an
Computer-Integrated Manufacturing, 23(3), 267-275.
Internet-enabled mobile device at any time.
[12] Alvares, A. J., & Ferreira, J. C. E. (2008). A system for
the design and manufacture of feature-based parts through
5 CONCLUDING REMARKS
the Internet. The International Journal of Advanced
In this paper, the concept of Cloud Manufacturing (CM)
Manufacturing Technology, 35(7), 646-664.
was presented. We analyzed the drivers, reviewed the current
[13] Sako, M. (2006). Outsourcing and offshoring:
status and predicted the future trends of CM. In addition, both
implications for productivity of business services. Oxford
short and long term impact of CM on three key sectors are
Review of Economic Policy, 22(4), 499-512.
discussed. As more research, development, and industry
[14] Hao, Q., Shen, W., & Wang, L. (2005). Towards a
applications are conducted, CM will enable manufacturing
cooperative distributed manufacturing management
enterprises to build virtual and open manufacturing processes,
framework. Computers in industry, 56(1), 71-84.
enabling its various stakeholders to connect and do business
[15] Opto 22. (2006). Understanding programmable
more seamlessly and cost effectively.
automation controllers (PACs) in industrial automation,
Web Link:

9 Copyright © 2013 by ASME

 http://www.newark.com/pdfs/techarticles/opto22/PACsInI [24] Parker, N. (2007). Intellectual property issues in joint
ndustrialAutomation.pdf ventures and collaborations. Journal of Intellectual
[16] Siemens. (2011). Web Link: Property Law & Practice, 2(11), 729-741.
http://www.automation.siemens.com/salesmaterial- [25] Huang, C. Y. (2002). Distributed manufacturing execution
as/brochure/en/brochure_totally_integrated_automation_o systems: A workflow perspective. Journal of Intelligent
verview_en.pdf Manufacturing, 13(6), 485-497.
[17] Panchal, J. H., & Schaefer, D. (2007). Towards achieving [26] Microsoft. (2011). Discrete manufacturing cloud
agility in web-based virtual enterprises: a decision-centric computing survey.
approach. International Journal of Internet Manufacturing [27] Viswanadham, N., & Gaonkar, R. S. (2003). Partner
and Services, 1(1), 51-74. selection and synchronized planning in dynamic
[18] Wagner, S. M., Eggert, A., & Lindemann, E. (2010). manufacturing networks. Robotics and Automation, IEEE
Creating and appropriating value in collaborative Transactions on, 19(1), 117-130.
relationships. Journal of business research, 63(8), 840- [28] Choy, K. L., Lee, W. B., Lau, H. C., & Choy, L. C.
848. (2005). A knowledge-based supplier intelligence retrieval
[19] Lomas, C. D. W. (2009). A design framework for agile system for outsource manufacturing. Knowledge-based
virtual enterprise collaboration (Doctoral dissertation, systems, 18(1), 1-17.
Durham University, UK). [29] Huang, X. G., Wong, Y. S., & Wang, J. G. (2004). A two-
[20] Torrisi, N. M., & de Oliveira, J. F. G. (2012). Remote stage manufacturing partner selection framework for
monitoring for high-speed CNC processes over public IP virtual enterprises. International Journal of Computer
networks using CyberOPC. The International Journal of Integrated Manufacturing, 17(4), 294-304.
Advanced Manufacturing Technology, 60(1), 191-200. [30] Choy, K. L., Lee, W. B., & Lo, V. (2002). An intelligent
[21] MTConnect. (2011). Web Link: supplier management tool for benchmarking suppliers in
http://mtconnect.org/media/7312/getting_started_with_mt outsource manufacturing. Expert Systems with
connect_-_final.pdf applications, 22(3), 213-224.
[22] OPC Foundation. (2012). Web Link: [31] Yoo, J. J. W., Kumara, S. R., & Simpson, T. W. (2012).
http://www.opcfoundation.org/Default.aspx/01_about/01_ Modular Product Design Using Cyberinfrastructure for
history.asp?MID=AboutOPC Global Manufacturing. Journal of Computing and
[23] Tao, F., Zhang, L., Venkatesh, V. C., Luo, Y., & Cheng, Information Science in Engineering.
Y. (2011). Cloud manufacturing: a computing and [32] Quirky. (2012). Web Link: http://www.quirky.com/
service-oriented manufacturing model. Proceedings of the [33] Ponoko. (2012). Web Link: https://www.ponoko.com/
Institution of Mechanical Engineers, Part B: Journal of
Engineering Manufacture, 225(10), 1969-1976.

10 Copyright © 2013 by ASME