Available online at www.sciencedirect.

com

ScienceDirect
Procedia CIRP 63 (2017) 39 45

The 50th CIRP Conference on Manufacturing Systems

Manufacturing system on the cloud: a case study on cloud-based process

planning
Xi Vincent Wang*, Mohammad Givehchi, Lihui Wang
Department of Production Engineering, KTH Royal Institute of Technology, Sweden

* Corresponding author. Tel.: +46-736-441-433; fax: +46-8-210-851. E-mail address: wangxi@kth.se

Abstract

The modern industry requires the next generation of manufacturing system to be intelligent, flexible, and interoperable. The manufacturing
system also needs to offer better performance with lower cost. The huge success of cloud computing brings the opportunity to the
manufacturing systems with high scalability, productivity and agility. It also provides practical solutions to the small and medium-sized
enterprises with elastic investment solutions. In a cloud manufacturing system, the manufacturing resource and capability are provided in the
service-oriented architecture as cloud services. In this research, the practical models of the cloud-based system are investigated, and distributed
process planning services are developed accordingly. The proposed infrastructure is evaluated via a case study in which the industrial
demonstrators are practised in the cloud as one specific type of process planning service.
2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
2017 The Authors. Published by Elsevier B.V.
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of The 50th CIRP Conference on Manufacturing Systems.
Peer-review under responsibility of the scientific committee of The 50th CIRP Conference on Manufacturing Systems
Keywords: Cloud, Manufacturing System, Process Planning;

robustness and sustainability of the manufacturing system.

1. Introduction
The trend of the modern production system is clear in
recent years, which is a collaborative environment supported
by Computer-Aided software (CAx) and networks.
Conventional paper-based technical documents have been
gradually replaced by the data files from computer-aided
software applications, e.g. Computer-Aided Design (CAD),
Computer-Aided Process Planning (CAPP), Computer Aided
Manufacturing (CAM), and Computer Numerical Controller
(CNC). These data files are exchanged over the network to
support the complex interactions between stakeholders like
multiple departments, providers and contractors. In this type
of environment, the manufacturers are required to provide
shorter responding time and higher agility to the fast changing
market and dynamic task assignments.
CAPP is a critical function bridging CAD, CAM and
Manufacturing Execution Systems (MES). At the middle
position of the so-called CAD-CAM-CNC chain, an advanced
CAPP system can provide adaptability, responsiveness,
Even though the CAPP system has been widely researched
and developed in both academia and industry, the
conventional CAPP system is incapable of meeting the
manufacturing trend mentioned above. The heterogeneous
data environment generated by multiple stakeholders and
dynamic contexts in modern manufacturing market put the
conventional CAPP system in the bottleneck position in the
middle of the CAD-CAM-CNC chain. Thus it is necessary to
improve the performance of current CAPP systems towards
the dynamic context between product design, tooling and
manufacturing companies. CAPP applications need to adopt
continuous technological improvement and innovation of
companies. Among multiple automation and information
technologies, Function Block (FB) and cloud show promising
potential to enhance the functionality and capability of CAPP
solutions. The rest of this paper is structured as follows.
Section 2 discusses the related research work on FB and cloud
technology. Section 3 proposes the cloud-bade process
planning system. Section 4 and 5 introduce the distributed
mechanism based on FB and its execution. The proposed

2212-8271 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of The 50th CIRP Conference on Manufacturing Systems
doi:10.1016/j.procir.2017.03.103
40 Xi Vincent Wang et al. / Procedia CIRP 63 (2017) 39 45
system is validated in section 6 via an industrial case study.
Finally, Section 8 concludes the paper with discussions.
2. Literature Review
2.1. Cloud-based system
In recent years, the Information and Communication
Technology (ICT) changed the landscape of the
manufacturing industry. Conventional paper-based technical
documents are replaced by data files processed computer-
aided design and manufacturing software applications.
Traditionally, the production work is centralised in the same
location or shopfloor close to the optimised resource or
logistic centre. Nowadays, the production departments,
providers and contractors can be distributed globally based on
their expertise and specific requirements. The product data
and document, e.g. process plans, can be easily exchanged
and transported over the network, which enables collaborative
and dynamic interactions between manufacturing
stakeholders. However, current CAPP systems are challenged
by the heterogeneous environment, which consists of multiple
CAD, CAPP, CAM and CNC applications that are developed
on different platforms using different programming languages
and communications protocols. Eventually, it is very difficult
for these applications to interface with each other due to
various interfaces and data formats. Among multiple
interoperable solutions, cloud technology is a promising
option to combat the challenge mentioned above.
The definition of cloud was initiated by NIST. In their
definition cloud is a computing model provides a shared pool
of configurable computing resources for ubiquitous,
convenient, on-demand network access [1]. These computing
services can be provided and released with minimal
management effort or service provider interaction. Then the
researchers in manufacturing discipline extend the computing
service to the manufacturing context. The concept of Cloud
Manufacturing (CM) is proposed to adopt cloud computing
technology in the manufacturing context and to improve the
current manufacturing systems [2-7]. In general, there are two
types of cloud manufacturing approach, which is directly
offering manufacturing software on the cloud as a service, and
providing manufacturing capabilities as an entirely new type
of cloud service [8].
The deployment of CM requires implementation of each
manufacturing application in the manufacturing cloud. In an
ASME (American Society of Mechanical Engineers) web
seminar, it was specifically pointed out that cloud providers
can support manufacturing applications with cheaper and
more scalable access [9]. Some industry has already
recognised the importance of the cloud approach. NVIDIA,
one of the largest Graphics Processing Units (GPU) providers,
has developed a cloud-based GPU that is capable of
delivering virtualised workstation performance and
capabilities to desktops [10]. It is predicted that cloud-based
CAD systems will be available in a very near future [11, 12].
In fact, Autodesk can already provide cloud-based
applications to support collaborative design [13]. Via the
environment called Autodesk 360, users can access some
simulations such as mechanical simulation for linear and
nonlinear static stress and dynamics, computational fluid
dynamics analysis for fluid flow and thermal simulations,
rendering, design optimisation, energy analysis, and structural
analysis.
From the cooperation perspective, Tai and Xu [14]
discussed the complexity of collaboration in a CM
environment. Cooperation is classified into four types, i.e.
random, operational, tactical and strategic. In a matured CM
environment, the system is required to process all types of
collaborative interactions at multiple levels, including typical
manufacturing applications such as simulation and
optimisation.
Chai et al. [15] established a High-Performance Cloud
Simulation Platform for efficient execution and problem
solving for complex simulation tasks. A simulation task is
developed as a cloud computing service. Computing
performance can be improved by the collaboration of
computing nodes in cloud computing approaches.
To facilitate a cloud-based on-line optimisation system,
Chandrasekaran et al. [16] developed a system named cloud
computing-based optimisation (CCBOD). In CCBOD, the
main server keeps the repository of data and carries out
optimisation. Its data input and retrieval mechanisms were
developed to cooperate with an online optimisation process.
However, the method of promoting an optimisation process in
the cloud is not as clear in the report, and this system is not
capable of identifying user/data in various domains.
2.2. Process Planning
Process planning in machining generally refers to those
preparatory tasks from design to manufacturing of a
mechanical product, such as process sequencing, machine and
cutter selection, tool path planning, operation optimisation,
NC code generation, as well as setup/fixture planning, etc.
Since the introduction of computers to the field of process
planning in the 1960s, subsequent research has been
numerous. By the end of 1980s, more than 156 computer-
aided process planning (CAPP) systems have been reported in
the literature survey by Alting and Zhang [17]. CAPP domain
is surveyed and reported by Marri et al. [18] in 1998. Xu et al.
[19] conducted a more recent survey on the related research,
including object-oriented approach, neural network-based
approach, Petri net-based approach, genetic algorithm-based
approach, multi-agent bidding-based approach, constraint-
based approach, feature-driven approach, and information and
knowledge management. The reported approaches in [19] and
their combinations have been applied to several specific
problem domains, such as setup planning, process sequencing,
tool selection, cutting parameter selection, and tool path
planning, to name a few. More recently, research efforts on
process planning have shifted to distributed process planning
[20], planning and scheduling integration [21], reconfigurable
process planning [22], and intelligent process planning based
on capacity profile of machine tools [23]. The common
objective of the recent research is to generate robust, precise
yet flexible process plans, effectively.
 Xi Vincent Wang et al. / Procedia CIRP 63 (2017) 39 45
Among proposed solutions, Cloud-DPP elaborates a two-
layer distributed adaptive process planning system based on
function-block technology [20]. FB-enabled intelligent
machine controllers are one of the key elements of Cloud-
DPP. This topic has been identified as one of the
methodologies that needs further investigation both for future
generation machine controllers with native FB support and for
legacy machines with traditional NC code support with
limited feedbacks. While no commercial FB-enabled
controller exists today, some research works have been
conducted in the area such as open CNC architecture based on
STEP-NC data model and IEC 61499 function blocks by
Minhat et al. [24]. Some other works related to FB controlling
of a CNC-milling machine and an industrial robot for
executing feature-based function blocks are reported in [25]
and [26].
Intelligent operation planning is another area that needs
further investigation according to the special needs in Cloud-
DPP. Intensive reasoning based on the features data of the
products models, machining knowledge, and resource data is
needed to be performed inside the FBs in machine controller
side. Machine controllers are considered intelligent units in
Cloud-DPP, which have the intelligence and know-hows
about how to manufacture the features. There is a long track
of research works in this area of feature-based machining and
knowledge-based operation planning until today. Using of
expert systems in manufacturing and process planning during
80s is surveyed by Gupta and Ghosh [27]. Carpenter et al.
[28] described methods and a computer based system for
automated machinability assessment and tool selection for
milling. Their research is based on the increasing demands on
competitive cutting tool manufacturers to supply a
comprehensive advice service with relation to selection of
appropriate tools and cutting data for different applications.
Arezoo et al. [29] presented a tool selection and condition
selection tool for machining operations using an expert
system. Denkena et al. [30] reviewed the CAPP systems from
the Knowledge Management perspective in 2007. Waiyagan
et al. [31] reported an intelligent feature-based process
planning for five-axis mill-turn parts that discussed about
machining knowledge representation as production rules of an
expert system. Park [32] covered a knowledge capturing
methodology in process planning. Li et al. [33] reported about
knowledge base and architecture of a STEP-NC Compliant
Intelligent Process Planning Module. A knowledge-based
hole-machining process planning system is proposed by Lee
et al. [34].
3. Cloud-based Process Planning
Despite the successful practice in CM and cloud-based
CAx systems, the current CAPP solutions are still not capable
of handling the dynamic context from the upcoming
disturbance, disruption and uncertainty during manufacturing,
e.g. breakdowns, dynamic changes of job priority, and
expected bottlenecks, etc. Thus it is necessary to develop a
series of innovative and adaptive cloud services in which
CAPP service works as the central service. In this research,
the proposed system adopts CM and service-oriented
41
computing approaches towards knowledge-based and efficient
process planning and simulation service. As illustrated in Fig.
1, the cloud service starts from the monitoring service for up-
to-date machine availability and utilisation to ensure that
decision making for planning and optimisation become
resource-aware and well informed. Then the knowledge-based
process planning is provided to distributed engineering teams
to create non-linear process plans collaboratively, using the
services of cloud manufacturing. Adaptive setup planning
tool generates setups adaptively based on the available
machines, their configurations and capabilities, by sharing
real-time in-process monitoring information from the
monitoring module. Eventually, the simulation service
provides an integrated process/setup simulation environment
to optimise parameters for reduced cycle time and resource
consumption
p and increased quality
q y and pproductivity.
y
Fig. 1. Process Planning-centric Cloud Manufacturing Service Model
(Adapted from [35])
The adaptive service proposed in the system is realised via
FB technology. The FB is the IEC standard [36] for industrial
measurement and controls for distributed system and devices.
The models and programmable control logic can be extended
to both hardware and software so as to integrate
manufacturing resources as cloud services. In general, the
planning algorithms and knowledge can be encapsulated
within FBs, and the system can combine and manipulate the
FBs via controlling the data and event flow through FBs.
In practice, there are different approaches of utilising FBs.
In this paper, three deployment models are summarised for
the Cloud-based Distributed Process Planning (Cloud-DPP)
service. In the first model (Fig. 2), the FBs are utilised at the
data level. The company hosts all the FBs for planning and
execution service internally. When necessary, the FBs are
capable of interfacing with manufacturing cloud and looking
for valuable planning knowledge, e.g. cutting tool
specifications and machining strategies, to support local
decision making without exposing any data to the cloud. This
model is specifically suitable for the companies with high
security and safety requirements. In this model, no intellectual
property is disclosed to the parties outside the company, while
the companies FB runtime system is supported by the rich
knowledge pool and library from the cloud.
In the second model (Fig. 3), the FBs are exchanged
among stakeholders as services. In some scenarios, one
company may not have all the required capability or
knowledge for a specific planning task. However, the
development of new FBs leads to long process and high costs.
42 Xi Vincent Wang et al. / Procedia CIRP 63 (2017) 39 45
So in that case, a company can download additional FBs from
the cloud FB library, which consists of a wide collection of
FBs uploaded by other companies. The FB download can be
defined as a leasing service with pre-defined cost and valid
period. When the downloaded FBs reach the end of the
service period, they expire and stop functioning. Eventually,
the FB solution of the company is a composition of both in-
house FBs and downloaded FBs from the manufacturing
cloud. Similarly to the first deployment module, all technical
details are maintained within the fences of the company,
which guarantees the safety and security of production data.
Fig. 2. Deployment Model 1: Data Exchange and Cloud Access
Fig. 3. Deployment Model 2: FB Downloading
In the third deployment model (Fig. 4), the FB network
reaches the advanced level, which is a dynamic runtime
system consisting of both local FBs and remote FBs executed
in the cloud. When a complex job is processed, the company
may need the support of both internal FBs and external FBs
from the cloud FB runtime. In that case, the internal FBs are
innitiated first. After excuted, they are connected to the
external FBs executed in the cloud. When needed, the data
and event triggers are sent to the cloud runtime over the
network, where the remote FBs are triggered and executed
afterwards. The execution results are then streamed back to
the local FB as data and event flows, which push the FB
network to the next stage. In that case, the executable FB
network is supported by both local FB system and the remote
cloud environment.
4. Distributed Process Planning Mechanism
A distributed process planning (DPP) architecture is
proposed with two main hierarchy levels in [Wang, et al.,
2003]. Supervisory planning is responsible for generating a
generic and machine-neutral process plan. A generic process
plan can be generated by grouping machining features such as
steps, slots, pockets, and holes into setups and sequences
according to their datum references, tool access directions and
some sequencing rules. Having access to the scheduling
results and machine tool data, the generic setups can be
further merged according to the machines specifications and
dispatched to CNC controllers by execution control. Finally
all the decisions regarding the tool and fixture selection, tool
path planning and tool parameter selection are made at the
operation planning level. Therefore a detailed process plan is
evolved by moving from the supervisory planning to the
operation planning level. FBs are used for encapsulation of
the data and process planning logic as well as execution of the
process during CNC machining. The proposed adaptive
process planning system can be positioned in a cyber-
workspace [37].
Fig. 4. Deployment Model 3: Cloud-based Function Block Network
5. Function Block execution on CNC machines
In supervisory planning phase of Cloud-DPP methodology,
through different steps of feature-model analysis, feature
sequencing, setup planning and machine assignment, a
generic process plan is generated. This process plan is in form
of a FB network and is to be deployed to the FB-enabled
controller of a chosen machine. Each machining feature
function block instance of the process plan program refers to a
corresponding FB type in an existing FB library in the
machine controller. Each machining feature function block in
the machine library contains the know-hows and ability to
provide the resource specific details of the process plan and to
cut the feature. These FBs are to be developed and embedded
by machine builders, operation planning experts, and
programmers. Therefore two units are required in the system.
One is a FB-enabled machine controller that is able to execute
a deployed function-block network corresponding to a
machining setup and the other is a development environment
to create machining feature FB library. When FBs for a
 Xi Vincent Wang et al. / Procedia CIRP 63 (2017) 39 45 43

certain set of machining features are developed and tested and host services to be instantiated and used in FB networks. Such
put into the Cloud-DPP library, that system is capable of distributed programs can be either generated
cutting parts including those features (i.e. what to do) without programmatically by FB System generation services on the
telling it how to do it. cloud or the web-based FB development environment GUI.
As it was mentioned before, the first requirement is having They are then stored in and accessible by FB systems
a FB-enabled machine controller. No commercial FB-enabled configuration host services. In Cloud-DPP the FB networks
controller exists today; however some research works are are mainly generated programmatically in an automated way.
conducted [24-26]. One solution is integrating a FB runtime Supervisory planning and execution control services call FB
environment hosted in a separate computer with a commercial system generation services in order to generate FB network
controller and considering the entire combination as the FB- representing the planned process. Generated FB network is
enabled controller (Fig. 5). The communication between the then deployed to the intended FB runtime environments to be
two elements is normally through Ethernet or serial executed. Targeted FB runtime environment here are mainly
communication cables. In cases when the machine controller the ones incorporated in FB enabled machine controllers.
has an operating system (OS) compatible with FB runtime
environment and Cloud-DPP HMI, and also allows
installation and some internal access to the resources, a better
solution exists. FB runtime environment and Cloud-DPP HMI
can be directly installed on the controller and the original
machine HMI is used with new GUI and functionalities. In
absence of APIs in controller that can be used to integrate on
OS level, an alternative solution is suggested in this research.
A virtual serial port emulator or a virtual file emulator can be
used to emulate a physical serial cable or an NC code file on
the file system. The communications are then possible to be
conducted through such virtual mediums.
Machine controllers might provide different levels of
access to its internal data and commands through a published
API (Application Programming Interface). To this extent, the
developed logical FB-enabled controller can have different
portions of functionalities that a native FB-enabled controller
provides. The minimum requirement that also works with a
large number of legacy controllers is the ability to send the
conventional NC code to the machine through DNC
Fig. 5. Integrating a commercial CNC controller and a FB runtime
functionality. The machine Control is supposed to be done by environment to be used as a FB-enabled machine controller in Cloud-DPP
FBs at runtime and therefore an adapting layer should be
created to convert the command to NC code at runtime. This 6. Case study
is in contrast with the approach of generating entire NC code
in advance and sending it to the machine to be executed with 6.1. KTH Manufacturing Cloud
no further control. NC commands in this way are generated at
runtime to translate the FB control intentions to the To evaluate the proposed system, a private cloud structure
commands understandable by the machine. Due to the fact is established with two rack servers, two CPU with 24 cores,
that planning and control is executed at runtime when the 64 GB RAM and 1 TB disk space. It is a reasonable size of
machining is being performed, operation plan adaptation and cloud facilities which is both affordable and effective for a
tuning is possible even during the machining process. It is manufacturing SME (Fig. 7). In this structure, Ubuntu Server
recommended to define a unique machine controller interface and OpenStack platforms are deployed to host the cloud
as an abstraction layer representing all machine tools in environment. The system can support various types of virtual
general to be used by feature FB to control different machines platform, e.g. Microsoft Windows-based and Linux-based.
in the same way. This unique machine control interface
should be then implemented and configured for different 6.2. CAPP-4-SMEs examples
machine types with suitable underlying implementations.
Regarding the second requirement on a development The developed platform is implemented and validated, in a
environment for making machining feature FBs for the real pilot case. During the systems operation, the monitoring
machines, a web-based development environment is service delivers all the required information regarding the
developed. Fig. 6 shows how FB types and FB control availability of the machine tools and their capabilities to the
programs are created, hosted, deployed, and executed in adaptive planning system.
Cloud-DPP. The web-based FB development environment Firstly, the cloud-based process planning service generates
provides the possibility to develop FB types for each a generic process plan by grouping machining features, such
machining feature type to be deployed to the FB libraries on as steps, slots, pockets and holes into setups and sequences.
the cloud. FB types can then be accessed through FB library Both setups and machining features are grouped on the basis
44 Xi Vincent Wang et al. / Procedia CIRP 63 (2017) 39 45

of datum references and manufacturing constraints. The
generated generic process plan is portable to different
machines having at least three axes. Once the generic process
plan has been generated, information related to the machine
monitoring is obtained in order to select not only the ones
available, but also the suitable machine tool to perform the
parts machining. In this case, the process planning service
selects the Mazak VTC-800/SR 5-axis machining center and
proceeds with adaptive setup planning. The optimised setup
plan is then generated based on the availability and the
capabilities of the selected machine tool.
determine the most suitable cutter available on the machine as
well as the optimal cutting parameters for each machining
operation. At the same time, the operation planning
algorithms make sure that all operations can finish within the
available time window of the chosen machine. The function
blocks are also used for machining execution by sending auto-
generated control codes to the machine controller. Fig. 9
shows the machined test part. Moreover, once the function
block network is deployed to the machine tool, the machining
of the test part is monitored and the status is available through
the monitoring services in real-time.

FB system Supervisory planning

generation and execution control
services services
ces

FB systems Monitoring
configuration services
host services
Fig. 8. FB network representing the generated process plan deployed to the
machine controller and being executed
FB library host
services

Web-based FB FB Runtime
Development Environment
Environment

Cloud-DPP GUIs Cloud-DPP HMIs

Fig. 9. Machined test parts
Fig. 6. Interaction of Cloud-DPP elements through which FB types and FB
control programs are created, hosted, deployed, and executed
7. Conclusions

A faster market clock speed of changes technological

innovation, which leads to co-evolution of products,
manufacturing processes and relevant systems [38], and
further demands efficient configuration and re-configuration
of manufacturing processes and systems to facilitate product
varieties and customisation [39]. However, conventional
CAPP applications are not capable of handling the dynamic
context from complex and dynamic data resources for multi-
criteria decision making and optimisation. Thus in this
research, the cloud and FB technologies are adopted in the
CAPP system thus forming the Cloud-DPP system. It
provides a promising capability of innovative and adaptive
process planning service via cloud and service-oriented
computing. Specifically, the event-driven FB technology is
Fig. 7. KTH Cloud Platform for Cloud-DPP
especially suitable for on-board adaptive decision making at
machine level to achieve real-time responsiveness,
In the current implementation, a process plan is
adaptability and overall resource effectiveness.
represented by a function block network with embedded
The executable FB system realises process plan
operation planning and execution algorithms. Fig. 8 shows the
embedding, smart algorithm invocation and bi-directional
FB network is deployed to the machine and is being executed.
communication at the shopfloor-level. As the output, the FB
By triggering the embedded algorithms, each FB can
network is successfully executed in the numerical controller
 Xi Vincent Wang et al. / Procedia CIRP 63 (2017) 39 45
during implementation and case study. Traditionally, the NC
machines are driven by the G-code based data files only. In
this research, adaptive and event-based FBs are successfully
developed in the cloud, and then deployed and executed in the
machine tool controller, which leads to the next generation of
machine controllers. The feasibility and adaptiveness of
Cloud-DPP are the main scientific contribution of the
proposed work. Conventionally, change of machines or
cutters leads to hours or even days of extra work. With the
help of the Cloud-DPP system, new process plans can be
generated in minutes. In the future, it is possible to extend the
FB development to more types of numerical controllers, and
open FB architecture can be one option of future CNC
developments.
References
[1] Mell, P.,Grance, T., 2011, The NIST Definition of Cloud Computing,
NIST Special Publication, 800/145: 7.
[2] Li, B., Zhang, L., Wang, S., Tao, F., Cao, J., Jiang, X., Song, X.,Chai, X.,
2010, Cloud Manufacturing: a New Service-oriented Networked
Manufacturing Model, Computer Integrated Manufacturing Systems,
16/1: 1-7.
[3] Xu, X., 2012, From Cloud Computing to Cloud Manufacturing, Robotics
and Computer-Integrated Manufacturing, 28/1: 75-86.
[4] Ren, L., Zhang, L., Tao, F., Zhao, C., Chai, X.,Zhao, X., 2015, Cloud
manufacturing: from concept to practice, Enterprise Information Systems,
9/2: 186-209.
[5] Tao, F., Zhang, L., Liu, Y., Cheng, Y., Wang, L.,Xu, X., 2015,
Manufacturing service management in cloud manufacturing: overview
and future research directions, Journal of Manufacturing Science and
Engineering, 137/4: 040912.
[6] Zhong, R.Y., Lan, S., Xu, C., Dai, Q.,Huang, G.Q., 2016, Visualization of
RFID-enabled shopfloor logistics Big Data in Cloud Manufacturing, The
International Journal of Advanced Manufacturing Technology, 84/1-4: 5-
16.
[7] Qu, T., Lei, S., Wang, Z., Nie, D., Chen, X.,Huang, G.Q., 2016, IoT-
based real-time production logistics synchronization system under smart
cloud manufacturing, The International Journal of Advanced
Manufacturing Technology, 84/1-4: 147-164.
[8] Wang, X.V.,Xu, X., 2014, Virtualize Manufacturing Capabilities in the
Cloud: Requirements, Architecture and Implementation, International
Journal of Manufacturing Research, 9/4: 348-368
[9] Knezevic, D., 2012, Changing the Way Engineers Work - Smartphones,
Tablets, the Cloud, and the Future of Engineering. ASME Webinar Series.
[10] NVIDIA, 2016, Deep Learning Solutions for Cloud Computing [cited
2017 Feb]; Available from: http://www.nvidia.com/object/gpu-cloud-
computing.html.
[11] Teschler, L., 2012, Graphics Processors Let CAD Go to the Cloud.
[cited 2012 December]; Available from:
http://machinedesign.com/article/graphics-processors-let-cad-go-to-the-
cloud-0531?page=0%2C0.
[12] NVIDEA, 2016, GRAPHICS-ACCELERATED VIRTUAL
DESKTOPS. [cited 2017 Feb]; Available from:
http://www.nvidia.com/object/nvidia-grid.html.
[13] Autodesk, 2017, Cloud Services: Design In The Cloud [cited 2017 Feb];
Available from: http://www.autodesk.com/360-cloud.
[14] Tai, D.Y.,Xu, F.Y., 2012, Cloud Manufacturing Based on Cooperative
Concept of SDN, Advanced Materials Research, 482: 2424-2429.
[15] Chai, X., Zhang, Z., Li, T., Zhang, Y.,Hou, B., 2011, High-performance
Cloud Simulation Platform Advanced Research of Cloud Simulation
Platform. in GCMS '11 Proceedings of the 2011 Grand Challenges on
Modeling and Simulation Conference. The Hague, Netherlands: Society
for Modeling & Simulation International.
[16] Chandrasekaran, M., Muralidhar, M.,Dixit, U., 2012, Online
Optimization of Multipass Machining based on Cloud Computing, The
International Journal of Advanced Manufacturing Technology, 65/1-4:
239-250.
45
[17] Alting, L.,Zhang, H., 1989, Computer aided process planning: the state-
of-the-art survey, The International Journal of Production Research, 27/4:
553-585.
[18] Marri, H., Gunasekaran, A.,Grieve, R., 1998, Computer-aided process
planning: a state of art, The International Journal of Advanced
Manufacturing Technology, 14/4: 261-268.
[19] Xu, X., Wang, L.,Newman, S.T., 2010, Computer-aided process
planning A critical review of recent developments and future trends,
International Journal of Computer Integrated Manufacturing, 24/1: 1-31.
[20] Wang, L.,Givehchi, M., 2011, Web-DPP : An adaptive approach to
planning and monitoring of job-shop machining operations, in
Proceedings of the 7th CIRP International Conference on Digital
Enterprise Technology. University of Patras.
[21] Ueda, K., Fujii, N.,Inoue, R., 2007, An Emergent Synthesis Approach to
Simultaneous Process Planning and Scheduling, CIRP Annals -
Manufacturing Technology, 56/1: 463-466.
[22] Azab, A.,ElMaraghy, H.A., 2007, Mathematical Modeling for
Reconfigurable Process Planning, CIRP Annals - Manufacturing
Technology, 56/1: 467-472.
[23] Newman, S.T.,Nassehi, A., 2009, Machine tool capability profile for
intelligent process planning, CIRP Annals - Manufacturing Technology,
58/1: 421-424.
[24] Minhat, M., Vyatkin, V., Xu, X., Wong, S.,Al-Bayaa, Z., 2009, A novel
open CNC architecture based on STEP-NC data model and IEC 61499
function blocks, Robotics and Computer-Integrated Manufacturing, 25/3:
560-569.
[25] Holm, M., Adamson, G., Wang, L.,Moore, P., 2012, An IEC 61499
Function Block based Approach for CNC Machining Operations, in
Proceedings of the 13th Mechatronics Forum. TRAUNER Verlag.
[26] Wang, L., Givehchi, M., Schmidt, B.,Adamson, G., 2012, Robotic
Assembly Planning and Control with Enhanced Adaptability, Procedia
CIRP, 3/0: 173-178.
[27] Gupta, T.,Ghosh, B.K., 1989, A survey of expert systems in
manufacturing and process planning, Computers in Industry, 11/2: 195-
204.
[28] Carpenter, I.D.,Maropoulos, P.G., 2000, A flexible tool selection
decision support system for milling operations, Journal of Materials
Processing Technology, 107/13: 143-152.
[29] Arezoo, B., Ridgway, K.,Al-Ahmari, A.M.A., 2000, Selection of cutting
tools and conditions of machining operations using an expert system,
Computers in Industry, 42/1: 43-58.
[30] Denkena, B., Shpitalni, M., Kowalski, P., Molcho, G.,Zipori, Y., 2007,
Knowledge Management in Process Planning, CIRP Annals -
Manufacturing Technology, 56/1: 175-180.
[31] Waiyagan, K.,Bohez, E.L.J., 2009, Intelligent feature based process
planning for five-axis mill-turn parts, Computers in Industry, 60/5: 296-
316.
[32] Park, S.C., 2003, Knowledge capturing methodology in process
planning, Computer-Aided Design, 35/12: 1109-1117.
[33] Li, P., Hu, T.,Zhang, C., 2011, STEP-NC Compliant Intelligent Process
Planning Module: Architecture and Knowledge Base, Procedia
Engineering, 15/0: 834-839.
[34] Lee, C.-S., Lee, J.-H., Kim, D.-S., Heo, E.-Y.,Kim, D.-W., 2013, A hole-
machining process planning system for marine engines, Journal of
Manufacturing Systems, 32/1: 114-123.
[35] CAPP-4-SMEs, 2013, CAPP-4-SMEs: Collaborative and Adaptive
Process Planning for Sustainable Manufacturing Environments. [cited
2013 May]; Available from: www.capp-4-smes.eu.
[36] IEC, 2005, IEC 61499. Function Blocks for Industrial-process
Measurement and Control Systems- Part 1: Architecture. Geneva: IEC.
[37] Wang, L., 2008, Wise-ShopFloor: an integrated approach for web-based
collaborative manufacturing, IEEE Transactions on Systems, Man, and
Cybernetics, Part C (Applications and Reviews), 38/4: 562-573.
[38] Tolio, T., Ceglarek, D., ElMaraghy, H., Fischer, A., Hu, S., Laperrire,
L., Newman, S.T.,Vncza, J., 2010, SPECIESCo-evolution of products,
processes and production systems, CIRP Annals-Manufacturing
Technology, 59/2: 672-693.
[39] Hu, S.J., Ko, J., Weyand, L., ElMaraghy, H., Lien, T., Koren, Y., Bley,
H., Chryssolouris, G., Nasr, N.,Shpitalni, M., 2011, Assembly system
design and operations for product variety, CIRP Annals-Manufacturing
Technology, 60/2: 715-733.