Computer Communications 151 (2020) 31–41

Contents lists available at ScienceDirect

Computer Communications
journal homepage: www.elsevier.com/locate/comcom

Intelligent manufacturing production line data monitoring system for

industrial internet of things
Wei Chen
School of Physics and Electrical Engineering, Weinan Normal University, Weinan 714000, Shaanxi Province, China

ARTICLE INFO ABSTRACT

Keywords: Applying the wireless sensor network of the Industrial Internet of Things and the radio frequency identification
Intelligent manufacturing technology to the production workshop of the discrete manufacturing industry, the real-time status of the shop
Industrial Internet of Things floor can be automatically collected, providing a powerful decision-making basis for the upper-level planning
Production line data monitoring
management department. This paper proposes a reference architecture and construction path for smart factories
by analyzing industrial IoT technology and its application in manufacturing workshops. Combined with the
analysis of the status quo and needs of the discrete manufacturing enterprise workshop, this paper designs
the overall architecture and theoretical model of the system. In view of the variety of on-site manufacturing
data, large amount of data, variable status, heterogeneity, and strong correlation between data, integrated key
technologies such as WSN and RFID, the industrial IoTs solution for manufacturing workshops is given. The
multi-thread data real-time collection, storage technology and product tracking monitoring of the workshop
are studied. Finally, the performance of the system is analyzed from the perspective of real-time and quality.
The results show that the system is effective in the monitoring of production line data.

1. Introduction competitiveness [5]. Intelligent manufacturing is the core of ‘‘Industry

4.0’’, and manufacturing IoT technology is the foundation of intelli-
The manufacturing workshop is the core of the company’s prod- gent manufacturing. It combines the technical means of IoT with the
uct production. The discrete manufacturing workshop is engaged in actual production of manufacturing to drive the production process
multi-variety and small-scale production, the production process is and change the traditional workshop management mode. Through the
complex, the production scheduling is difficult, and the monitoring identification of work-in-progress, dynamic tracking of the in-process
and management of the discrete manufacturing workshop has always product and real-time collection of production data can be realized,
been a problem that plagues the enterprise [1]. In the context of eco- real-time detection of status information can be realized, and trans-
nomic globalization, market competition has expanded from regional parency of the manufacturing process of the shop can be improved [6].
to global, and manufacturing is facing severe pressure to survive [2]. In addition, the real-time positioning technology of the workshop has
At the same time, customers’ individualized demands are increasing, attracted more and more attention from the manufacturing enterprises.
the uncertainty of survival orders is further enhanced, how to respond Through the real-time positioning of the production factors of the work-
to diversified demands in a timely manner, adjusting the production
shop, the position and state of each production factor are understood,
process, shortening the production cycle and ensuring product qual-
the workshop distribution process is optimized, the object search time is
ity, is the survival of enterprises. And the primary issue that needs
reduced, and the comprehensive object of the workshop is realized [7].
to be addressed. However, opportunities and challenges coexist. The
Although the research on smart factories has just started, the defi-
development of informatization and networking has brought tremen-
nitions and ideas of smart factories have been proposed from different
dous impact and change to traditional business ideas and management
angles at home and abroad. Relevant scholars have made a specific
methods, and also pointed out the direction for the development of
definition of the smart factory [8,9]. They believe that the smart factory
manufacturing enterprises [3]. Traditional order-oriented production
methods are gradually shifting to service-oriented production meth- is based on the manufacturing of the object, through the data analysis
ods [4]. Through information technology, enterprises can realize real- to find the factory operating rules, using the rules to achieve intelligent
time control of the production process of the workshop, timely adjust decision-making, and then package the intelligent decision-making into
production plans according to market changes or customer needs, intelligent services. Through the cloud agile configuration to achieve
improve market response speed, optimize manufacturing resources, service synergy, a new product of the factory is formed by self-learning
improve production efficiency and reduce costs, and improve core and self-adaptation [10]. Manufacturing IoT is a new manufacturing

E-mail address: chenwei@wnu.edu.cn.

https://doi.org/10.1016/j.comcom.2019.12.035
Received 31 October 2019; Received in revised form 4 December 2019; Accepted 19 December 2019
Available online 20 December 2019
0140-3664/© 2019 Elsevier B.V. All rights reserved.
W. Chen Computer Communications 151 (2020) 31–41

mode that combines IoT technology with manufacturing technology and explained. On this basis, the overall industrial IoT system of the
to realize the process of product manufacturing and service, as well intelligent workshop of the system was designed. The communication
as the dynamic perception, intelligent processing and optimal control of communication terminals such as intelligent workshop CNC machine
of manufacturing resources and information resources throughout the tools, intelligent workshop wireless sensor network and intelligent
product life cycle [11,12]. Some scholars have studied the application workshop radio frequency identification network are designed.
of RFID technology in the resource allocation of clothing manufacturing The rest of this article is organized as follows. Section 2 discusses
industry [13,14]. Through RFID real-time data, using fuzzy theory to the industrial IoT technology for intelligent manufacturing, and Sec-
deal with inaccurate information, combined with the characteristics of tion 3 studies the overall design of the production line data monitoring
clothing manufacturing industry, they proposed a resource allocation system. Section 4 analyzes the functions of the workshop IoT and
system based on RFID to realize the optimization of manufacturing production line data monitoring system. Section 5 discusses the results
resource allocation. The application results in a clothing enterprise in- of the production line real-time data monitoring system, and Section 6
dicates that the system can optimize resource allocation [15]. Through summarizes the full text and gives the future research direction.
the electronic identification of the item-level objects, the comprehen-
sive collection of the quality data of the underlying items is realized, 2. Research on industrial IoT technology for intelligent manufac-
and the adaptive method of knowledge learning is used to realize turing
the quality control of the manufacturing process [16]. Aiming at the
dynamic production process of enterprises, some scholars put forward 2.1. Smart factory RFID technology
an RFID-based enterprise application integration framework, and gave
a method to realize production dynamic management and work-in- RFID technology is a non-contact automatic identification technol-
process visualization under the framework, and verified the feasibility ogy that requires no human intervention. It consists of electronic tags,
and reliability of the system [17,18]. In order to improve the efficiency antennas, readers and application layer software. The electronic tag
of workshop monitoring, the researchers proposed a model-based work- is bound to the item, and can store a small amount of coded data,
shop monitoring method, built a state compilation model through RFID generally adopting the methods of adhesion, printing, slot, etc., which
observation variables, monitored production processes and anomalies is equivalent to the carrier of the product, and distributed among
based on event analysis and event processing techniques, and intro- the manufacturing enterprises, the market, and the user. The tool
duced Hidden Markov method for state diagnosis and prediction of the for encoding data is connected to the transmission network and the
production process [19,20]. Related scholars studied the complex event client, and the related information is entered and updated through the
construction method based on RFID raw data, defined the event by application and the database, and the corresponding information of the
binary matrix, and considered the timing relationship between events, product is obtained. The system database can be accessed through the
established a state monitoring model, realized the detection of inter- Internet anywhere in the world, thereby achieving tracking of items and
ference events, and carried out case analysis, verifying the feasibility remote information query and management to achieve a global physical
and effectiveness of monitoring [21,22]. Siemens designed a ‘‘digital interconnection.
factory’’ platform to divide the production process of the plant into The reader sends RF signals continuously at a certain frequency.
several production nodes, track the manufacturing process information When the electronic tag enters the read/write range of the reader, the
and data of each production node through real-time monitoring, and electronic tag receives the RF signal sent by the reader and obtains
integrate and analyze the data for production. Process optimization, energy through the induced current to issue the encoded informa-
equipment fault diagnosis, MES (manufacturing execution system) and tion stored in the tag chip. Read by the reader, the reader receives
supply chain management have achieved the purpose of information and decodes the encoded information sent by the electronic tag, and
interconnection from the lower floor production to the upper planning then sends the decoded information to the PC application via the
management department. USB/RJ45/RS232/WIFI interface for corresponding processing.
This paper proposes a reference architecture for smart factories by RFID technology can read multiple tags at the same time, that is,
studying the application of IoT technology in manufacturing work- multiple objects can be identified and read at the same time, which
shops. In view of the shortcomings of the chaotic and robustness of most is suitable for situations where multiple entities share resources. The
common manufacturing workshops, the extraction and monitoring of information storage capacity of the electronic tag chip is much larger
manufacturing process information is single and lagging, the system than the previous bar code and can be set to read and write passwords,
cost is high, and the upgrade and maintenance are difficult. The real- and the security is high.
time tracking and monitoring system of intelligent workshop products Choosing the right operating frequency is a very important step
based on the Internet of Things is proposed. The principle of IoT in RFID technology. Current RFID operating frequencies are broadly
related technology and its implementation method in the workshop are classified into low frequency systems, high frequency systems, and
analyzed. The related concepts of the existing smart factory are deeply microwave systems. The data storage capacity, size, and read/write
analyzed, and the reference architecture and construction path of the distance of the reader are also different for the different working fre-
smart factory are proposed. The industrial IoT system is the basis of quencies. The higher the working frequency, the farther the recognition
the smart factory. The intelligent manufacturing workshop is a core and reading distance is, the faster the speed, the larger the data storage
component of the smart factory. Through the analysis of the status quo capacity of the tag, and of course the higher the price. Since the power
and demand of the workshops of discrete manufacturing enterprises loss of the low frequency is proportional to the cube of the propagation
in a certain area, the overall architecture and theoretical model of distance, the power loss of the high frequency is proportional to the
the system are designed, including the system functional structure square of the propagation distance, so the high frequency can also be
model, system business process model and system architecture model. used for tracking and positioning the label.
We designed and built the manufacturing workshop IoT, including the In the workshop, the state of the product undergoes a step-by-step
networking of equipment such as CNC machine tools in the workshop, process, the state is constantly changing, the position is constantly
wireless sensor network, radio frequency identification network, wire- changing, and there is a large amount of dynamic information. There-
less control network and network integration, and proposed the method fore, the application of RFID technology to the workshop, the electronic
and principle of real-time data acquisition monitoring and product tag and the shop product are bound, not only can store the static
tracking in the workshop, providing hardware network foundation and information of the product in the whole life cycle, but also can read
technical theory support for the design of system software. The concept and write the real-time dynamic information of the product through
of the industrial Internet of things in the workshop was analyzed the RFID reader without interruption. The physical topology of the

32
W. Chen Computer Communications 151 (2020) 31–41

Fig. 1. Real-time monitoring system physical topology based on discrete manufacturing process of industrial Internet of Things.

discrete manufacturing workshop real-time monitoring system based on as well as the operating status and parameters of the equipment are the
the industrial Internet of Things is shown in Fig. 1. factors that affect the quality of the workshop products. By applying
The system sets corresponding RFID readers and electronic bill- WSN technology to the workshop, not only can the sensor node idle,
boards in each station in the workshop. The RFID readers collect work status, intelligent scheduling, and dispatching of the sensor nodes
real-time data in the production process. The current production tasks be used to maximize the utilization of the equipment; it can also collect
and production status can be transmitted to the workers through the the displacement, temperature, and speed of the equipment during
electronic signage of the station. Each RFID reader is bound to the the manufacturing process in real time. The main operating parameter
corresponding station, and sets the processed area and the area to be data that affects the processing quality, such as vibration, adjusts the
processed. When the production object with the electronic label enters operating parameters of the equipment in real time to ensure the
the corresponding area, the data can be collected and according to the quality of the product, so that the production equipment is always in
reader. The bound logical area gets its current location information. the optimal energy efficiency state. According to the data collected for
a long time, the coupling relationship between the various influencing
2.2. Smart factory WSN technology factors during the operation of the equipment can be analyzed, and the
variation rules of the operating parameters of the equipment can be
The wireless sensor network integrates the sensor node module, the summarized, thereby predicting the abnormal trend of the equipment,
data conversion processing module and the data wireless transmission monitoring the equipment health condition and early warning of the
communication module, and is composed of a sensor interconnec-
fault.
tion system capable of not only data acquisition and processing but
also wireless communication transmission. Through the cooperation of
2.3. Intelligent factory based on industrial internet of things
multiple integrated sensor nodes, the information to be tested in the
monitoring area is collected and transmitted, thereby completing the
collection, processing and analysis of information at any time in any The smart factory is in the ‘‘Internet +’’ environment, facing the
place in the monitoring area. development needs of digital, network and intelligent manufacturing.
The wireless sensor network architecture generally consists of sen- Through the integration of the Internet field and industrial manufactur-
sor nodes, aggregation nodes, and task management nodes. The sensor ing, we will promote the innovation of the manufacturing development
nodes are further divided into an anchor node and an unknown node model and the construction of an ecological modern industrial system.
according to the known location. The location known node is called an Germany’s ‘‘Industry 4.0’’ and the ‘‘Industrial Interconnection’’ pro-
anchor node, and the location of the unknown node can be determined posed by the United States are typical forms of smart factories. The
indirectly according to the distance between the anchor node and the reference architecture of the smart factory proposed in this paper is
anchor node. roughly summarized as the following four aspects.
The workshop has the characteristics of complex environment and (1) Collaborative design and manufacturing process simulation op-
many disturbances, and many environmental factors of the workshop, timization

33
W. Chen Computer Communications 151 (2020) 31–41

Based on cloud platform and big data driven resource scheduling by the machine, and the interference of many factors such as dust
and information design of service design, collaborative design can be and electromagnetic field, it will have certain influence on the data
realized. The simulation of manufacturing process based on virtual transmission, especially for collecting digital signals and analog signals.
simulation technology is based on the relevant data of the whole life In terms of frequency, it should have high anti-interference to overcome
cycle of the product, modeling, simulation and control optimization the interference factors in the harsh environment of the workshop, and
of all production process flow such as workshop layout, equipment ensure accurate and effective data collection and transmission.
and manufacturing process. It builds complex industrial system models (2) Real-time
that integrate knowledge, data and models, and coordinates production Real-time is the fundamental guarantee for enterprise production
and management elements, and links multi-domain and multi-level decision-making and control. If the real-time nature of the manufac-
knowledge. Based on real-time data production system parameters and turing process data of the equipment under test, product tracking
state identification, it is a process simulation system for intelligent and positioning data, etc. cannot be guaranteed, then the authenticity
control of manufacturing equipment and production process. of the data at the determined moment is lost, and the control and
(2) Industrial Internet of things system scheduling of the workshop is meaningless, which will result in a series
Industrial IoT system is an indispensable part of intelligent factory of problems. Moreover, due to the large number of monitoring units and
to achieve data collection, condition monitoring and information trans- parameters, and all parameters need to be displayed in real time, it is
mission, and provides support for intelligent monitoring and control of necessary to use multi-threading to ensure the real-time performance
CPS manufacturing process. The industrial internet network system is of all data.
shown in Fig. 2. (3) Rapid deployment capability
(3) Intelligent manufacturing workshop Due to the large number of equipment types in the workshop and
The intelligent manufacturing workshop is a highly intelligent work- the different operating systems, there is a demand for cross-platform
shop based on the industrial Internet of Things, which automatically and compatible integration capabilities of the system to achieve rapid
tracks and monitors the basic elements of workshop products and deployment and cost savings.
equipment, and intelligently controls the manufacturing process. The (4) Convenient expansion
intelligent control of the manufacturing process automatically performs As enterprises are constantly developing, their business and needs
the decision-making process based on the decision-making knowledge are not always the same, which requires the system to be easily ex-
base and manufacturing data by real-time storage extraction, analysis panded and upgraded to adapt to the business changes and evolving
and processing of various manufacturing data such as tooling, process, needs of the enterprise.
and workpiece, as well as production equipment status and runtime (5) Security
parameters. Security is a problem that must be considered by any industry. The
Intelligent warehouse management is based on DFID, and real- main security issues considered by this system are to prevent external
izes electronic and automated warehouses through electronic tags and attacks from causing data leakage or tampering, and on the other hand,
stackers, making materials entering and leaving the warehouse, ware- data or program loss caused by hardware device failure or improper
house scheduling, and inventory management highly intelligent. operation by the workshop. Data storage exceptions are caused when
(4) Supply chain and operation service management parallel data volumes are overloaded, which imposes requirements
Aiming at the needs of manufacturing supply chain resource integra- on the system’s architecture, security policies, and ability to handle
tion, efficient management and service model innovation, a cloud-based exceptions.
supply value chain collaboration system supporting network manufac-
turing is established. Based on multi-system data fusion, through big 3.2. System function structure model
data analysis and mining technology and artificial intelligence tech-
nology, the computer software automatically proposes optimization The overall functional structure model of the system is divided into
opinions on the operation, management and decision-making of the six modules: system login, DNC, workshop monitoring, warehouse man-
enterprise, and can obtain optimized results based on the comparison agement, product tracking, and statistical analysis. Each main function
of the data before and after optimization. module includes multiple sub-function modules, as follows:
(1) System login module
3. Research on the overall design of the production line data The module mainly includes four sub-function modules: user reg-
monitoring system istration, user login, password management and rights management,
to realize the security protection of the system, and open different
3.1. System requirements analysis corresponding modules for users with different positions in different
departments to view the modification authority and data security for
For the needs of enterprise management and actual production the enterprise.
process, from the perspective of technical theory, it can be summarized (2) DNC module
into the following three functional requirements: (1) The networking The module mainly includes five sub-function modules: process
of the production equipment, touch screen and other terminals in the document management, NC equipment management, NC code trans-
workshop realizes real-time information transmission with the upper mission, drawing model transmission, and process card transmission,
department; (2) Workshop equipment status, manufacturing process so as to establish serial port and network port communication with the
information and warehousing information are monitored in real time communication machine such as CNC machine tools, equipment and
and presented in the management planning department with an intu- touch screens in the workshop. The device sends and receives process
itive software interface; (3) Product parts, standard parts, tooling, etc., documents such as NC codes, process cards, 2D drawings, 3D models,
should be monitored in real time. and the management of these documents.
Therefore, based on the above functional requirements, an intel- (3) Workshop monitoring module
ligent workshop production line data monitoring system based on The module is mainly for the monitoring of each workshop, real-
industrial Internet of Things is designed, and the system needs to have time extraction of the detailed manufacturing process data such as
the following performance: the running status, operating parameters, current machining parts,
(1) Anti-interference production task completion of each workshop, and real-time adjust-
Due to the harsh environment in the workshop, the complicated ment equipment operating parameters, monitoring equipment health
lines, the complicated production process, the noise and heat generated status, and according to this, it carries out on-demand distribution

34
W. Chen Computer Communications 151 (2020) 31–41

Fig. 2. Industrial Internet of Things network architecture.

of production tasks, fault diagnosis, maintenance alarms and other collected and stored in the remote database server for the whole system
actions. call.
(4) Warehouse management module
These factors are taken into account: (1) The harsh environment
The module mainly includes four sub-function modules: RFID device
management, storage management, outbound management, and inven- of the workshop and the complexity of the lines will have a certain
tory management, to realize RFID reader parameter and mode setting, impact on network cabling and data signal transmission. Therefore, the
electronic label initialization, product storage coding and information wireless connection is used for the whole workshop network, and the
entry. AP client access point client mode is selected. Under the premise of
(5) Product tracking module ensuring signal strength, the wireless AP is used to extend the coverage
The module mainly includes two sub-function modules: product of the network to the entire workshop to realize wireless transmis-
real-time status tracking and product real-time status query. The track-
sion of the network; (2) Considering the security of the network and
ing and query information includes: basic product information, current
data, all the workshops of the Internet of Things are integrated into
location information of the product, current process information of
the product, current processing equipment information of the product, the workshop Ethernet, and then connected to the corporate Internet
current processing progress of the product, and current responsible through the firewall; (3) Each terminal device is coded by IP address
information of the product. in the workshop LAN, and each terminal device is bound with a fixed
(6) Statistical analysis module IP address to achieve accurate command and data transmission.
The module mainly includes: equipment utilization statistics, equip-
The system architecture model is shown in Fig. 3. In the workshop,
ment failure rate statistics, product qualification rate statistics, product
the Internet of Things first needs to connect the communication ter-
failure reason statistics, and workshop production statistics. It generates
statistical analysis charts or reports to provide a reliable data basis minals such as CNC machine tools and touch screens in the workshop
for management analysis departments to make reasonable planning to realize real-time communication between the upper design manage-
decisions, rewards and punishments, and equipment maintenance. ment department and the workshop, transmission of NC code to CNC
machine tools, transmission of process documents, and release of tasks.
3.3. System architecture model Secondly, the sensors and RFID devices arranged in the workshop need
to be networked, and the real-time data, positioning information and
The client/server mode is the mainstream choice of the current
real-time data of the workshop production are converted into wireless
computing mode. The client passes the data to the server. The server an-
network collection data acquisition host through the sensors and RFID
alyzes the data passed and performs related operations on the database
according to the analysis result, and then returns the result to the client. readers arranged in the workshop, and then through the workshop.
The C/S mode has the characteristics of strong interactivity, safe data Ethernet uploads and saves data to the real-time database server for
storage mode, short response time, low communication volume, and office area retrieval and visual real-time monitoring. Then, it is the
good processing of large amounts of data. Due to the large amount networking of the execution terminal of the workshop. The workshop
of data interaction in this system, the data needs multi-thread real- control center makes management and scheduling decisions according
time acquisition, and requires high real-time interactive responsiveness to the real-time production data of the workshop, and sends control
and data security. Therefore, the classic C/S architecture system is
commands to the integrated controller through the wireless network to
adopted here, and the computer system and workshop are connected
control the intelligent execution units of the workshop. Finally, all the
through the workshop Internet of Things. CNC equipment such as
communication terminals, RFID readers, sensors and other hardware networks are integrated into the workshop Ethernet, and the enterprise
devices are connected, and the corresponding client is installed on Internet is connected through the firewall to ensure network and data
each terminal client of the workshop, and real-time data information is security.

35
W. Chen Computer Communications 151 (2020) 31–41

Fig. 3. System architecture model.

4. Workshop internet of things and production line data monitor- of the limitation of the power access point, so that the AP The wireless
ing system function research access device is directly pasted on the CNC machine tool, and then
wirelessly connected with other APs through the AP client access point
4.1. Networking design of CNC machine tools and touch screens in the client mode of the wireless access device, and all the machine tools and
workshop touch screens in the workshop are set to a fixed IP address and a MAK
address.
In order to realize the real-time transmission of manufacturing data In the local area network between the established CNC machine
to the workshop, the NC code is transmitted to the CNC machine tool and the server, the communication and file interaction are realized
tool, and the process card, 2D drawing, 3D model, etc. are transmitted by accessing the shared files under the respective IP addresses. The
to the station touch screen. The working status of the CNC machine network module of the machine tool is similar to the working mode of
tool and the production data are uploaded to the office area in real the PC network card, and consists of hardware and software. Hardware
time. However, the brands and models of CNC machine tools used by refers to the network card and network card driver, and the software
various enterprises are different. Due to the updating and introduction is the network communication module of the machine tool.
of machine tools, there are various types of CNC machine tools in
different workshops, which leads to inconsistencies in the reserved 4.2. Design and function realization of intelligent workshop wireless sensor
network interfaces. At present, the reserved networking interface of network
CNC machine tools in China is generally RS232, RS485 serial interface
and RJ45 Ethernet port. Therefore, for the above serial interface and (1) Design of workshop wireless sensor network
network port, the networking scheme of workshop CNC machine tool In view of the complex environment and high interference of the
and station touch screen is designed as shown in Fig. 4. workshop, the wireless sensor network of the workshop needs to ar-
Due to the complex environment in the workshop, the numerous range the sensor nodes reasonably, and collect the real-time working
lines, and the distance from the office area, the wire line should be state of the equipment through the sensor to collect the signal light of
minimized, the line anti-interference ability should be enhanced, and the equipment, the magnetic flux of the motor or directly collect by the
the serial transmission distance should be overcome. Here, the RS232 PLC, through different sensors.
and RS485 serial ports are converted into RJ45 Ethernet ports through The sensor nodes arranged in the workshop continuously transmit
the serial port server, and then connected to the wireless access device the digital and analog signals to the coordinator. The coordinator
to be converted into a wireless network. The wireless connector is transmits the signals to the processor for AD conversion and signal
powered by the Passive Po E network cable, so that the AP can get rid amplification through the serial port asynchronous communication,

36
W. Chen Computer Communications 151 (2020) 31–41

4.3. Design and function realization of the network in the workshop radio
frequency

(1) Design of workshop radio frequency identification network

The intelligent workshop radio frequency identification network
is established, so that all entities in the workshop have their own
unique identification, and their attributes, processes and locations can
be queried at any time. This provides great convenience to the manage-
ment. On the one hand, it can track and query the processing progress
of WIP at any time. The time required for each process is stagnation,
and the cause is found in time and intelligently dispatched. On the other
hand, it can be tracked, avoiding the problems of manual registration in
traditional workshops, difficulty in finding tools and tooling. The smart
factory radio frequency identification network reference architecture is
shown in Fig. 6.
The architecture generally follows the IoT general architecture sys-
tem of the sensing layer, the network layer, and the application layer.
The bottom-up is an RFID tag that corresponds to the entity object, and
the static data is stored, and the RFID reader perceives the entity object.
(2) Intelligent production workshop production line data monitor-
ing function design
The communication process of the RFID application system is that
the reader/writer module is connected with the host computer via
the USB/RJ45/RS232/WIFI interface, and receives and executes the
commands sent by the host computer one by one, and finally returns
the executed result information to the upper computer.
When the host computer sends the command data block to the
Fig. 4. Intelligent workshop CNC machine tool and workstation touch screen RFID reader module, the interval between adjacent characters must be
networking solution. less than 12 ms. Otherwise, the previously sent data will be directly
discarded by the RFID reader module, and the data after 12 ms will
be treated as a new one. The command data block is re-received until
and then converts the signals into wireless signals through the wireless the read–write module receives the correct command, executes the
network card. The wireless monitoring video signal transmitted by command, and returns a response to the reader module.
the high-definition camera is sent to the wireless sensing gateway The process of sending the response data by the reader/writer is
together, and finally enters the workshop local area network and is actually the process of returning the execution result and the response
stored in the database server, and the application terminal such as the data to the upper computer after the reader receives and executes the
control center converts the collected voltage, current and other signals command sent by the upper computer, which is a complete communica-
into corresponding collection parameter values through a calculation tion process. The memory of the electronic tag can be logically divided
program. The design of the workshop wireless sensor network reference into four memory areas, each of which typically contains one or more
layout scheme is shown in Fig. 5. memory words. The four storage areas are an EPC area, a TID area, a
(2) Design of workshop status monitoring function user area, and a reserved area, respectively.
The wireless sensor network communicates with many sensor nodes Each product is attached to the warehouse with an electronic label,
arranged in the monitoring area and self-organizes through the aggre- and the product code and other information is written into the elec-
tronic label storage area, and the relevant information is entered into
gation node, and finally transmits the collected data information to the
the database. The product code finds the unique product information of
remote task management node for related processing. The sensor node
the database. By arranging the corresponding readers and configuring
usually consists of the following four parts:
the client program at the appropriate position in the workshop, the
The sensor module is used to collect various information of the
principle of ‘‘one read, one end and multiple bits’’ is followed. When the
object to be collected in the monitoring area and convert the collected
product passes through the reader, it will be automatically recognized,
data into a format convenient for transmission. The processor module is
and the current location, current state, time and other information will
used to store and process the data collected by the sensor module. The
be written into the electronic tag storage area according to the fixed
wireless communication module is configured to convert the processed
encoding format or the database data will be updated in real time
data signal into a wireless signal, and exchange control information according to the number, as shown in Fig. 7, thus realizing the real-time
with other sensor nodes, and send and receive data. The energy supply tracking query for the workshop product.
module is used to energize the sensor nodes to enable them to function
properly. 5. Production line real-time data monitoring system results anal-
After the above sensor nodes are powered on, the parameters are ysis
initialized, the input signal is converted by AD, the corresponding
channel is selected, and the data information collected by the sensor 5.1. Real-time analysis
is sent to the coordinator according to the set frequency, and then
transmitted to the processor for processing according to the asyn- This section analyzes and studies the operation results of product-
chronous serial communication protocol. After that, it is converted into based data acquisition and monitoring systems from the aspects of
a wireless signal through a wireless network card, and is connected hardware and software execution efficiency and production efficiency
to the workshop LAN and the Wi Fi/GPRS network, and finally the before and after system implementation.
collected manufacturing data is displayed in the smart phone terminal On the hardware, Line Server runs on high-performance servers,
and the control center. and its processing speed is higher than that of ordinary computers or

37
W. Chen Computer Communications 151 (2020) 31–41

Fig. 5. Workshop wireless sensor network design architecture.

Fig. 6. Smart factory manufacturing workshop radio frequency identification network design architecture.

industrial computers. In terms of data and real-time signal processing, By counting the running time markers of the PLC, the machining
its speed can meet the requirements of industrial sites. In software cycle of each station includes an online machining cycle and an offline
design, Line Server uses a linked list to manage the connection with machining cycle. The online machining cycle process includes: the
each device, and scans the common monitoring bits of each station workpiece reaches the fixture position, the feedback of the confirmation
every 480 ms. In the database query speed, the database is installed signal, the device startup processing, the upload of the processing data,
locally. In terms of network, since the PLC and the Line Server are and the device reset. There are no data acquisition and control signals
directly connected through the Ethernet and in the same LAN segment, for offline processing.
the speed of the network transmission is related to the performance of The time consuming of the data acquisition system and the control
the switch, and the transmission delay through the system tester is less system is the difference between the time taken by the line processing
than 90 μs. cycle and the time taken by the offline processing cycle. Table 1 records

38
W. Chen Computer Communications 151 (2020) 31–41

Table 1
Average processing time of online and offline for each station in each month.
Station Online Offline Functional Production line Production line
(OP No) processing processing average time MES function MES function
cycle (s) cycle (s) consuming (s) average time overall time
consuming (s) consuming (s)
12 14.10 13.10 0.80 – –
21 12.50 11.40 0.70 – –
32 9.40 8.50 0.50 – –
40 8.20 7.20 0.80 – –
54 10.10 9.10 0.70 – –
61 16.90 16.30 0.80 – –
72 6.80 6.20 0.90 0.42 5.00
83 0.90 0.80 0 – –
94 0 0 0 – –
102 19.70 18.80 0.60 – –
113 0.08 0.10 0 – –
121 14.70 13.70 0.10 – –
134 20.20 20.10 0.10 – –

Fig. 7. Using RFID technology to implement product monitoring methods.

the online and offline processing time statistics for a certain month. The
time-consuming fluctuations of different stations are relatively large,
which is related to the process complexity of different stations. Here,
the relative size of the system time-consuming and offline processing
time is mainly analyzed. Fig. 8 shows the online machining cycle,
offline machining cycle, and average time consumption. The longer the
line passes through the x-axis, the shorter the time it takes for the data
acquisition and control system.
After statistical analysis, the average time-consuming data acquisi-
tion and monitoring system of the single line of the production line is
0.42 s, which can meet the real-time requirements of the automation
site. The entire production line completes one workpiece. The average
time taken after implementing the system is about 127.6 s. The exe-
cution time of data acquisition and control is 5.4 s, which accounts
for 4.10% of the processing time of the entire production line. It has
basically no effect on the completion of the production plan. Fig. 8. Online and offline processing cycle diagram of each station in a month.

5.2. Quality control analysis

efficiency of production lines. The equipment will verify its state before
Key indicators of production process control of enterprises KPIs
include rework rate, defective rate, yield rate, scrap rate and other indi- producing a single workpiece, and will remind the workpiece that is
cators, which is an indicator to evaluate the comprehensive production not qualified or produced according to the process requirements, and

39
W. Chen Computer Communications 151 (2020) 31–41

Table 2 platform is engaged in leasing services, and each company rents its own
Statistics of rework/unqualified times of each station in a month.
modules on the cloud platform according to their respective needs and
Station (OP No) Unqualified (times) applies them to actual enterprises.
12 486
21 0
Declaration of competing interest
32 8
40 12
54 2 The authors declare that they have no known competing finan-
61 0 cial interests or personal relationships that could have appeared to
72 5 influence the work reported in this paper.
83 64
94 64
102 152 CRediT authorship contribution statement
113 6
121 42 Wei Chen: Conceptualization, Methodology, Writing - original
134 8
draft, Formal analysis.

Acknowledgments
stop production, thus eliminating the non-conforming product and the
workpiece that does not follow the process flow into the next process. This work is supported by the Shaanxi Military and Civilian Integra-
The rework data table counts the number of rework per station. tion Research Fund Project, China, Research of PID Controller Applied
The production process of this production line is complicated, the to Pubai Xigu Coal Mine Gas Drainage System (18JMR40); Project of
work station is more, the process is crossed, and the rework operations Weinan Normal University, China, Research on Tuning Methods of PID
of different stations are different. For example, the OP90 is pre-installed Controller Parameters used in Process Control System (7YKS06).
for the bracket, there is no rework operation, and rework is not possi-
ble. The pre-installation failure is directly scrapped, and the OP70 is References
reworked. By configuring in the configuration interface, the workpiece
[1] B. Yun, Industrial internet of things over tactile internet in the context of
will automatically reach the rework station after entering the produc- intelligent manufacturing, Cluster Comput. 21 (4) (2017) 1–9.
tion line again. Other stations will skip or prohibit production, the [2] R.Y. Zhong, X. Chen, C. Chao, et al., Big Data Analytics for Physical Internet-
unconfigured workpiece can only be reworked in the current process. based intelligent manufacturing shop floors, Int. J. Prod. Res. 55 (9) (2017)
2610–2621.
The number of times of rework can also be limited according to the
[3] P. Lade, R. Ghosh, S. Srinivasan, Manufacturing analytics and industrial internet
relevant configuration to ensure the orderly production and process of things, IEEE Intell. Syst. 32 (3) (2017) 74–79.
controllability. [4] Y. Lv, D. Lin, Design an intelligent real-time operation planning system in
Table 2 counts the number of unqualified stations in a certain distributed manufacturing network, Ind. Manage. Data Syst. 117 (4) (2017)
742–753.
month. The total output of the production line in the month is 31,217,
[5] H. Panetto, P.C. Stadzisz, W. Li, et al., Guest editorial: Special issue on
of which 30,752 is qualified, the number of workpiece rework is 177, (industrial) internet-of-things for smart and sensing systems: Issues, trends, and
and the number of processing scraps is 916. In the process of scrapping, applications, IEEE Internet Things J. 5 (6) (2019) 4392–4395.
the terminal inspection was 142 times, accounting for 14.15% of the [6] M. Roopaei, P. Rad, K.K.R. Choo, Cloud of things in smart agriculture: Intelligent
irrigation monitoring by thermal imaging, IEEE Cloud Comput. 4 (1) (2017)
total scrap rate. The scrap rate for the month was 2.79%. When the
10–15.
data collection and control system was not deployed, the average scrap [7] G. Heo, J. Jeon, A study on the data compression technology-based Intelli-
rate was 3.22%, the end-of-line scrap rate was 74.39%, and the scrap gent Data Acquisition (IDAQ) system for structural health monitoring of civil
rate was reduced by about 18%. The ratio of scrapped at the terminal structures, Sensors 17 (7) (2017) 1620.
[8] P. Patel, M.I. Ali, A. Sheth, From raw data to smart manufacturing: AI and
was reduced. 76%. semantic web of things for industry 4.0, IEEE Intell. Syst. 33 (4) (2018) 79–86.
[9] Sedat Bingöl, Hidir Yanki Kiliçgedik, Application of gene expression program-
ming in hot metal forming for intelligent manufacturing, Neural Comput. Appl.
6. Conclusion
30 (3) (2018) 937–945.
[10] J. Qian, B. Zi, D. Wang, et al., The design and development of an omni-
This paper analyzes the related concepts of the existing smart directional mobile robot oriented to an intelligent manufacturing system, Sensors
17 (9) (2017) 2073.
factory and proposes the reference architecture and construction path
[11] B. Martinez, M. Montón, I. Vilajosana, et al., Early scavenger dimensioning in
of the smart factory. The industrial IoT system is the foundation of wireless industrial monitoring applications, IEEE Internet Things J. 3 (2) (2017)
smart factories, and the intelligent manufacturing workshops make the 170–178.
core components of smart factories. Through the analysis of the related [12] Farshid Farnood Ahmadi, Integration of industrial videogrammetry and artificial
neural networks for monitoring and modeling the deformation or displacement
technologies of IoT and its implementation methods in the workshop,
of structures, Neural Comput. Appl. 28 (12) (2017) 3709–3716.
the core technologies of two IoT technologies, such as radio frequency [13] Z. Tao, H. Zhang, C. Zhu, et al., Design and operation of App-based intelli-
identification and wireless sensor networks, are applied to design the gent landslide monitoring system: the case of Three Gorges Reservoir Region,
industrial Internet of the workshop. The real-time tracking of the Geomatics Nat. Hazards Risk 10 (1) (2019) 1209–1226.
[14] M. Babar, F. Khan, W. Iqbal, et al., A secured data management scheme for
manufacturing workshop products and the real-time monitoring of the
smart societies in industrial internet of things environment, IEEE Access 6 (99)
workshop status and the workshop equipment status are realized, which (2018) 1.
proves the correctness and feasibility of the research theory and key [15] A. Singh, A. Gaur, A. Kumar, et al., Sensing technologies for monitoring
technologies. However, real-time monitoring of the manufacturing pro- intelligent buildings: A review, IEEE Sens. J. 18 (12) (2018) 4847–4860.
[16] J. Penn, P. Pennerstorfer, A. Jungbauer, New generation of continuous casting
cess is only the first step toward intelligent manufacturing. To achieve
plants with intelligent manufacturing strategy, Berg- Huettenmaenn. Mon.hefte
intelligent production, production factors need to have the ability of 163 (1) (2018) 11–17.
self-identification, intelligent perception, independent decision-making [17] X. Chen, S. Zhang, J.M.P. Geraedts, Guest editorial focused section on sensing
and knowledge learning. This requires further exploration in the areas and perception Systems for Intelligent Manufacturing (SPIM), IEEE/ASME Trans.
Mechatronics 23 (3) (2018) 983–985.
of manufacturing IoT technology, artificial intelligence algorithms, and
[18] Y. Chen, G.M. Lee, L. Shu, et al., Industrial internet of things-based collaborative
machine learning. In the future, the system can be combined with cloud sensing intelligence: Framework and research challenges, Sensors 16 (2) (2016)
service technology to hang each functional module in Yunping. The 215.

40
W. Chen Computer Communications 151 (2020) 31–41

[19] L.M. Camarinha-Matos, S. Tomic, P. Graça, Technological innovation for the Wei Chen was born in Henan, China, in 1984. She re-
internet of things, IFIP Adv. Inf. Commun. Technol. 25 (2) (2016) 617–622. ceived the BS degree from China University of Mining &
[20] D. Mishra, A. Gunasekaran, S.J. Childe, et al., Vision, applications and future Technology, Xuzhou, China, in 2006, and the ME degree
challenges of Internet of Things, Ind. Manage. Data Syst. 116 (7) (2017) from China University of Mining & Technology, Beijing,
1331–1355. China in 2010. After graduation she worked as a teacher
[21] P. Fragalamas, T.M. Fernándezcaramés, M. Suárezalbela, et al., A review on in School of Physics and Electrical Engineering at Weinan
internet of things for defense and public safety, Sensors 16 (10) (2016) 1644. Normal University since 2010. She has got a further study
[22] K. Kaur, S. Garg, G.S. Aujla, et al., Edge computing in the industrial internet of in Shaanxi Normal University in 2017. She has published 12
things environment: Software-defined-networks-based edge-cloud interplay, IEEE papers and holds 3 patents. Her research interests include
Commun. Mag. 56 (2) (2018) 44–51. wireless sensor networks, and advanced control theory.
Email: chenwei@wnu.edu.cn

41