ArticlePDF Available

Abstract and Figures

In the era of market globalisation, the quality of products has become a key factor for success in the manufacturing industry. The growing demand for customised products requires a corresponding adjustment of processes, leading to frequent and necessary changes in production control. Quality inspection has been historically used by the manufacturing industry to detect defects before customer delivery of the end product. However, traditional quality methods, such as quality inspection, suffer from large limitations in highly customised small batch production. Frameworks for quality inspection have been proposed in the current literature. Nevertheless, full exploitation of the Industry 4.0 context for quality inspection purpose remains an open field. Vice-versa, for quality inspection to be suitable for Industry 4.0, it needs to become fast, accurate, reliable, flexible, and holistic. This paper addresses these challenges by developing a multi-layer quality inspection framework built on previous research on quality inspection in the realm of Industry 4.0. In the proposed framework, the quality inspection system consists of (a) the work-piece to be inspected, (b) the measurement instrument , (c) the actuator that manipulates the measurement instrument and possibly the work-piece, (d) an intelligent control system , and (e) a cloud-connected database to the previous resources; that interact with each other in five different layers, i.e., resources , actions , and data in both the cyber and physical world. The framework is built on the assumption that data (used and collected) need to be validated, holistic and on-line, i.e., when needed, for the system to effectively decide upon conformity to surpass the presented challenges. Future research will focus on implementing and validating the proposed framework in an industrial case study.
Content may be subject to copyright.
Multi-Layer Quality Inspection System Framework for Industry 4.0
Paper:
Multi-Layer Quality Inspection System Framework
for Industry 4.0
Victor Azamfirei, Anna Granlund, and Yvonne Lagrosen
alardalen University
15 Hamngatan, Eskilstuna 632 20, Sweden
Corresponding author, E-mail: victor.azamfirei@mdh.se
[Received February 23, 2021; accepted April 26, 2021]
In the era of market globalisation, the quality of
products has become a key factor for success in the
manufacturing industry. The growing demand for
customised products requires a corresponding adjust-
ment of processes, leading to frequent and necessary
changes in production control. Quality inspection
has been historically used by the manufacturing
industry to detect defects before customer delivery
of the end product. However, traditional quality
methods, such as quality inspection, suffer from
large limitations in highly customised small batch
production. Frameworks for quality inspection have
been proposed in the current literature. Nevertheless,
full exploitation of the Industry 4.0 context for quality
inspection purpose remains an open field. Vice-versa,
for quality inspection to be suitable for Industry 4.0,
it needs to become fast, accurate, reliable, flexible,
and holistic. This paper addresses these challenges by
developing a multi-layer quality inspection framework
built on previous research on quality inspection in the
realm of Industry 4.0. In the proposed framework, the
quality inspection system consists of (a) the work-piece
to be inspected, (b) the measurement instrument,
(c) the actuator that manipulates the measurement
instrument and possibly the work-piece, (d) an
intelligent control system, and (e) a cloud-connected
database to the previous resources; that interact with
each other in five different layers, i.e., resources,
actions,anddata in both the cyber and physical
world. The framework is built on the assumption
that data (used and collected) need to be validated,
holistic and on-line, i.e., when needed, for the system
to effectively decide upon conformity to surpass the
presented challenges. Future research will focus on
implementing and validating the proposed framework
in an industrial case study.
Keywords: quality inspection, Industry 4.0,
cyber-physical systems, zero-defect manufacturing,
CAD/CAM/CAE
1. Introduction
In the era of customer-oriented products, the manufac-
turing industry is facing unprecedented challenges. In-
dividualised products require a corresponding adjustment
of processes, leading to frequent and necessary changes in
production control. These changes must be implemented
efficiently so that companies survive the changing market
requirements [1].
Further, the quality of products should still be
considered to secure the gain of customers and market
share [2]. Product quality deviation can lead to numerous
problems, such as a high rate of re-work or scrap, inferior
product functional performance, tooling failures, and
unexpected production downtime, thus, reducing both
production and product quality [3]. Quality inspection is
usually adopted by manufacturing companies to capture
deviations before customer delivery. This involves
activities such as “measuring, examining, testing, or
gauging one or more characteristics of the product and
comparing the results with specified requirements” [4].
Nevertheless, traditional quality methods, such as
quality inspection, suffer from significant limitations
in highly customised small-batch production [5]. For
instance, exploiting the increasingly available data and
adapting to dynamic changes[6]. In [1], it is pointed that
quality inspection needs to become: “fast,” “accurate,
“reliable,” “flexible,” and “holistic” in order to be suitable
for Industry 4.0. The last two being directly related to
the individualisation of products and the related dynamic
changes in production. In addition, the productivity,
performance, and quality of products are affected by
the conditions of machines, manufacturing processes,
and manufacturing decision-making [2]. For production
to meet zero defects, symbiosis of information and
communications technology (ICT), artificial intelligence
(AI) models, quality inspection tools [7], and data usage
to adapt to dynamic environments as changing factors [8]
is needed. Therefore, modern production systems have
increasingly integrated new key enabling technologies
(KET), such as in-line data gathering solutions, data
storage, communication standards, data analytic tools,
and digital manufacturing technologies [6].
In the academic literature, data fusion for analysis
of product geometry quality has been studied [9],
Int. J. of Automation Technology Vol.15 No.5, 2021 641
https://doi.org/10.20965/ijat.2021.p0641
© Fuji Technology Press Ltd. Creative Commons CC BY-ND: This is an Open Access article distributed under the terms of
the Creative Commons Attribution-NoDerivatives 4.0 International License (http://creativecommons.org/licenses/by-nd/4.0/).
Azamfirei, V., Granlund, A., and Lagrosen, Y.
as well as estimation of part quality in multistage
manufacturing [10]. Frameworks for quality inspection
have been proposed in [11–15]. Nevertheless, full
exploitation of the different dimensions that constitute
Industry 4.0 for quality inspection is briefly discussed
in [16, 17]. To our knowledge, a holistic picture of
the different dimensions of quality inspection and their
interaction, e.g.,
The symbiosis of ICT with AI models or cyber
world, with
The physical world,
is missing. The purpose of this paper is to present an
overview of the latest advances in the quality inspection
field in regard to Industry 4.0 and develop a framework
for quality inspection for its full exploitation in the
Industry 4.0 context. The paper combines research
within cyber-physical systems (CPS), AI, and cloud
computing, allowing for quality inspection in customised
production. Future research challenges of the proposed
quality inspection framework are discussed in the context
of Industry 4.0 and mass customisation.
This paper is structured as follows. In Section 2,
the method for selecting relevant literature for this study
is explained. In Section 3, the literature review of
quality inspection frameworks and trends in the expansion
(dimensions and variables) is presented. In Section 4, a
detailed description of the functionality and information
needed for enhancing quality inspection are elaborated. In
Section 5, the proposed framework is described. Finally,
Section 6 concludes the paper and discusses future work.
2. Research Method
In order to carry out this literature review, the
authors benefited from Scopus and Google Scholar as
search engines, using mainly (but not exclusively) four
publishers: IEEE, Emerald Insight, Taylor & Francis, and
Science Direct. Google Scholar was used to covering
related publications by snowballing the references from
the selected articles. The search has been limited to:
The manufacturing industry and Industry 4.0,
The subject areas of Engineering and Computer
Science,
Document type, i.e., article, conference paper,
review, and book chapter,
Written in English, and
Published after 2011 when Industry 4.0 had, for the
first time, been introduced in Germany [1].
Additionally, standards [18–20] and well-established
literature in quality inspection, such as [21, 22], have been
used.
The relevance of search results to the subject of study
was judged as suggested in [23], i.e.,
Table 1. Search term groups and the publications reviewed
per group.
Search term (group) Publications
reviewed
Quality Inspection (general) 5
QI +
Industry 4.0 12
Artificial intelligence 23
Cloud computing 2
Cyber-physical systems 8
Digital twin 3
Intelligent robotics 6
Zero defects manufacturing 4
In-line 9
Total reviewed (excluding duplicates) 61
Browsing the title and abstract, as well as,
Introduction and conclusions,
Speed-reading to identify the most important parts
for the research, and
Reading in detail those identified parts.
This literature review was conducted by first searching
the term quality inspection and understanding the
meaning and usage in the manufacturing context. The
next natural step was to overview the progress of quality
inspection – and the activities that encompasses as
describedinSection1–inIndustry 4.0, highlighting
the different solutions for enhancing quality inspection.
Note that the concept of Industry 4.0 is a collective term
that encompasses many modern automation systems, data
exchanges, and production technologies. Industry 4.0
components were originally thought of as CPS,smart
factories, and smart products. Topics such as zero defect
manufacturing,Internet of Things (IoT),andintelligent
robotics were also included since they have a direct
impact on quality. These headings were analysed in detail
within the scope of their relation to quality inspection. See
Tabl e 1 for the resulting number of reviewed publications.
3. Literature Review
Extensive research has been dedicated to the quality
inspection field; nonetheless, the available guidelines for
performing quality inspection, such as ISO 17020 [24],
have been left to company interpretation. Quality
inspection is used to determine the product’s conformance
to the manufacturing standards, involving activities such
as measuring, examining, and testing of characteristics.
For this purpose, seven essential (but not limited) steps of
the fundamentals of quality inspection are proposed [21]:
Interpretation of specification,
Measurement of the quality of the characteristic,
642 Int. J. of Automation Technology Vol.15 No.5, 2021
Multi-Layer Quality Inspection System Framework for Industry 4.0
Fig. 1. Inspired by Juran’s seven essential steps for quality
inspection procedure [21].
Comparison between (i) “interpretation of specifica-
tion” and (ii) “measurement,
Judgment on conformance,
Process of conforming items,
Disposition of nonconforming items, and
Record of obtained data;
see Fig. 1.
In Juran’s model, the information flow starts with
the actual specifications created by the designer. The
interpretation of the specifications and its respective
assumptions affect the overall quality of the inspection.
Finally, recordings of the obtained data would be used
to update the specifications if needed; thus, having a
closed-loop model; see Fig. 1.
Even though Juran’s model is suitable for both manual
and automatic inspection, it was highly influenced by
human behaviour. Favourably, since the human inspector
is part of a socio-technological system, he/she is aware
of the manufacturing process trend, the people involved
in the manufacturing process, and their work quality.
This serves as an advantage for the inspector since it
provides assumptions of possible defects location and,
hence, a predefined inspection strategy. On the contrary,
the human inspector’s interpretation of the standard might
imply a certain degree of deviation from the intended
specifications. Later, statistical process control (SPC)
has been integrated into the quality inspection realm to
the evaluation of product characteristics, allowing for
a process-near control loop and fast correction of the
production process before faulty parts are produced [22].
Measurement system analysis is a set of guidelines for
assessing the quality of a measurement system issued
by the Automotive Industry Action Group (AIAG) [18].
For measurement system analysis (MSA), measurement
system is defined as “the collection of instruments
or gauges, standards, operations, methods, fixtures,
software, personnel, environment, and assumptions used
to quantify a unit of measure or fix assessment to
the feature characteristic being measured; the complete
process used to obtain measurements” [18]. In order for
Fig. 2. Inspired by AIAG and Pendrill’s measurement
system [18, 25].
it to perform correctly, the measuring process needs to
be free of uncertainty. The acronym SWIPE (Standard,
Work-piece, Instrument, Person and Procedure, and
Environment) is used to represent the six essential
elements of a generalised measuring system to assure
the attainment of the required objectives. One external
interpretation of the “measurement system” is defined in
Pendrill [25] in his Measuring Man framework. This is
interpreted in Fig. 2,inwhich:
The object is measured by
Ameasuring instrument or the operator’s own
sensing,
Being the operator that manipulates the measuring
instrument and needs to make the decision upon
conformance by also considering
The measurement method,and
The environment.
In the measurement system model, the measurement
information is transmitted from the measurement object,
often via an instrument, to an operator. The object,
instrument, and operator are the main elements of the
measurement system, but the measurement method or
environment can affect the main system elements when
determining overall measurement quality; see Fig. 2.
Another helpful model for defining a measurement
system by its primary sources of variation is PISMOEA
(Part, Instrument, Standard, Method, Operator, Envi-
ronment and Assumptions), which supports universal
application [18].
From the standardisation of quality inspection (e.g.,
MSA) described above, different studies have been done
in contribution to the field. Germani et al. [26] developed
an integrated inspection system that enables the designer
to define the tolerances and supervise the inspection
process. Jaramillo et al. [27] expanded the previously
described work and proposed a real-time CAD-based
inspection of deformable parts, in which they eliminated
the need for work-piece pre-alignment using a virtual
fixturing process to generate virtual deformations on the
CAD model.
In the realm of zero defect manufacturing and
Industry 4.0 – in which manufacturing industry and
information technology are linked together – Imkamp
et al. [1] forecasted the future five development topics of
Int. J. of Automation Technology Vol.15 No.5, 2021 643
Azamfirei, V., Granlund, A., and Lagrosen, Y.
quality inspection to address the Industry 4.0 challenges
and trends:
Fast: reduce measuring time.
Accurate: reduce measurement errors and cope with
uncertainties.
Reliable: verification of measurement uncertainties.
Flexible: increasing information density and variety
of measurement techniques.
Holistic: the relevant quality characteristics are
brought together to form a complete basis for
evaluating product quality.
From these development topics, the first three are mostly
related to technological advances, while the last two also
involve a methodology paradigm change.
With the increasing integration of non-destructive mea-
surement techniques, the level of complexity increases
when assessing conformance [28]. Molleda et al. [28]
discussed a decision support system for in-line assessment
of welds quality utilising statistical analysis of different
manufacturing variables as well as previously historical
data gathered from similar welds. Within this model,
“preventive maintenance” and “control” are constantly
updated. Wang et al. [29] combined a structured light
system with data mining and RFID technology for quality
inspection. Their quality inspection system is composed
of four levels, which are
Measurement level,
Data process level comprises quality-related feature
determination and extraction,
Computational intelligence level uses data mining
approaches to achieve the automated quality
classification based on the feature vector, and
RFID hardware and software for traceability.
Takaya [16] reviewed technical trends in in-process and
on-machine measurements for quality management and
control of products. In his work, Takaya suggested
an expansion of quality inspection from the assessment
of geometrical features/properties to generating function
of products, i.e., a comprehensive/strategic closed-loop
control; thereby, integrating CAD/CAM/CAE with
holistic measurement techniques. Li et al. [17] developed
an integrated feature-based dynamic control system
for on-line machining, inspection, and monitoring.
In their system, quality inspection is triggered by
machine monitoring when encountering product or
process deviations. The quality inspection system will,
thereby, be planned for the tool path compensation
automatically based on dynamic feature information to
overcut machining errors.
Stroppa et al. [30] proposed a multi-agent system in
which different quality control agents – that perform
in-line quality inspection – interact between themselves.
Thereby, it creates a high level of knowledge that permits
the system to self-optimise/reconfigure itself in terms of
hardware and software. This allows it to react to changes,
both expected (based on the results of past inspections)
and unexpected (e.g., decide if all the predefined tests are
necessary or if some of them can be skipped). Kiraci
et al. [31] studied the effects of measurement accuracy
and repeatability for robot measurement systems. They
considered insufficient the usage of the mainstream SPC
during the introduction of new products to identify and
eliminate defects. This was based on the assumption
that SPC and similar tools do not prevent defects from
occurring/recurring; thus, Kiraci et al. required the
development of intelligent control systems to mitigate
the variation of manufacturing processes. S ¨oderberg
et al. [32] presented the necessary inspection data for
real-time optimisation of geometry assurance in full
production using a digital twin. Furthermore, S¨oderberg
et al. procedure recommended the link of inspection data
with metadata, describing the assembly process and part
manufacturing status. Majstorovic et al. [33] in line
with the demands of smart factory (i.e., optimisation,
flexibility, self-adaptability and learning, fault tolerance,
and risk management), developed the framework for
cyber-physical manufacturing metrology model. This
framework consists of the following submodules:
Recognition of the CAD model geometrical features,
Intelligent inspection planning,
Measuring instrument cloud-connected, and
Analysis of results and generation of reports
available for all interested parties in the product
lifecycle.
More recently, Gao et al. [34] emphasised the
importance of in-process metrology in their literature
review. They described how the system benefits
from quality inspection for machining compensation,
i.e., feedback from the surface measurement data is
received into the machine controller to modify the tool
path, thereby reducing the machining deviations. Babu
et al. [3] developed a Spatio-Temporal Adaptive Sampling
methodology for estimation of whole part deviations
based on partial, historical, and simulated measurements.
The necessity of partial measurements arises due to the
high cycle time of 3D-surface scanners compared to the
assembly system cycle time. Syam et al. [35] presented
and demonstrated a methodology for the development of
fast and effective in-line optical measuring instruments
for the surfaces of parts. Based on the usage of any
available information that can improve the measurement
process, the proposed methodology uses three phases:
A priori knowledge and data gathering,
Instrument and software development, and
Usage of in-line instrument for on-line control.
644 Int. J. of Automation Technology Vol.15 No.5, 2021
Multi-Layer Quality Inspection System Framework for Industry 4.0
They claimed that data fusion methods that allow
the synthesis of inspection, monitoring, and database
data (CAD/CAM/CAE) are essential to produce more
consistent, accurate, and useful information. Phan
et al. [36] proposed a scan path/trajectory planning
method for on-machine measurement in a 5-axis machine
tool. The originality of their work lies in the fact that
the trajectory allows not only to take scanning quality
constraints into account but also to minimise the time
allocated to the measurement operation.
Xu et al. [37] proposed an industrial internet
architecture applicable to quality inspection in which edge
computing and cloud computing are used to collect data
and make decisions through AI with high performance
and low processing latency. Schmitt et al. [15]
presented a holistic approach for predictive model-based
quality inspection using machine learning and edge cloud
computing.
Zhang et al. [38] discussed standardisation directions
for machine vision and developed a machine vision
on-line detection system framework for intelligent man-
ufacturing that meets the standardisation requirements.
Evangelista et al. [14] presented the SPIRIT software
framework for robot quality inspection that combines two
sub-frameworks:
“Off-line sub-framework” that uses model-based
automatic coverage planning for complex parts and
generates a robot program, and
“In-line sub-framework” that executes the inspection
task on the robot and deals with sensor data mapping
for transferring sensor measurements to the 3D
object model.
Finally, Ding et al. [39] overviewed AI-powered
manufacturing and its applications in monitoring tasks,
such as quality inspection. AI technologies bring
novel manufacturing modes that possess intelligent
characteristics, such as self-perception, self-comparison,
self-prediction, self-adaptation, and self-optimisation.
From this literature review, it can be concluded that
quality inspection is still an open field of research,
spreading in different directions, i.e., technological
and methodological. The growing available holistic
data demand interoperability (e.g., cloud computing),
intelligent modules (e.g., AI), and a methodological
change for quality inspection. In Section 4, different
trends for enhancing quality inspection identified in the
literature are structured and interpreted. Following, in
Section 5, a framework that encompasses the different
trends is presented.
4. Analysis of Constituting Parts of Quality
Inspection
To adapt quality inspection to individualised produc-
tion, i.e., mass customisation, ICT plays an essential
role in the interaction between value-added process
(e.g., machining or assembly), monitoring system, and
inspection system. Three different identified trends for
enhancing quality inspection are described in this section:
Holistic data,
Intelligent control system, and
Cloud-connected resources.
The interpretations on how the quality inspection should
be enhanced are later represented in Fig. 3 and explained
in Section 5.
4.1. Holistic Data
In Juran’s model [21], the quality inspector is part
of a higher, more complex socio-technological system
and, thereby, influenced by external information, e.g.,
other operators. Juran describes it as the advantage to
“estimate” where defects are most likely to occur and,
thereby, dedicate extra attention. Consequently, this
would involve ignoring the least likely defects that might
occur. For Juran’s model, that is when quality inspection
starts, with the actual knowledge of the operator on the
product and production. On the contrary of having a
human in the loop, the human inspector’s interpretation
of the standard might imply a certain degree of deviation
from the intended specifications. This has been later
expressed as part of quality inspection in the PISMOEA
model under the “assumptions” [18].
Nevertheless, the initial intention of quality inspection
is to decide upon conformity of the manufacturing
process in regards to the designer’s intention. Thereby,
Germani et al. [26] decided to enable the designer
to define the tolerances and supervise the inspection
process. In addition, Phan et al. [36] considered
additional constraints to the quality inspection, such
as “measurement quality” (higher accuracy implies an
increase in measuring time). We interpret inspection
constraints as any additional restriction on the quality
inspection system, e.g., measuring quality, time available
for inspection and re-inspection, and additional features
to be inspected.
The manufacturing process inevitably affects the
product’s geometrical features as well as functionality.
Li et al. [17] proposed the feature-based dynamic
control triggered by the actual machine performing the
value-added operation. Further, Takaya [16] proposed
the integration of CAD/CAM/CAE. In addition, historical
data on the system’s performance should be recorded for
aiding the decision-making [3,28] and avoid historical
problems from re-occurring. Stroppa et al. [30] created
a higher level of knowledge that permits the inspection
machine to self-optimise by enabling integration between
different quality control agents. This is connected with
Imkamp et al. [1] need for holistic evaluation of the
system. Fig. 4, below, is our interpretation of how a
database should perform within the context of quality
inspection:
Int. J. of Automation Technology Vol.15 No.5, 2021 645
Azamfirei, V., Granlund, A., and Lagrosen, Y.
Fig. 3. Representation of the proposed multi-layer quality inspection framework for Industry 4.0.
Fig. 4. Holistic data for quality inspection.
Insertable data in which the designer or production
engineer add additional inspection constraints to the
system such as time to inspect.
Database or available data in which (a) specifica-
tions (i.e., desired outcome from the manufacturing
system) for geometric dimensioning and tolerancing
(GD&T) (CAD) as well as manufacturing (CAM)
and engineering (CAE), (b) monitoring data of
the value-added machines prior to inspection, and
(c) historical measuring data.
Holistic extracted data from the quality inspection
system.
The control system should benefit from the
database in performing pre-measurement activities
in which deviations are estimated based on quality
inspection and production systems’ performance
and, consequently, selecting or generating the
discrete measuring strategy.
4.2. Intelligent Control System
With the increasing amount of available information,
strategic inspection is important to avoid time-consuming
Fig. 5. Required intelligence for a rich data quality inspection.
measurement procedures [3, 16]. Recent work in the
field has introduced the usage of data mining and
analysis for quality inspection [29]. Kiraci et al. [31]
required the development of intelligent control systems
to mitigate the variation of manufacturing processes and
prevent historical defects from re-occurring. AI models
possess intelligent characteristics, such as self-perception,
self-comparison, self-prediction, self-adaptation, and
self-optimisation [39], which make them perfect for
quality inspection as control system.
Nonetheless, when Imkamp et al. [1] forecasted the
usage of holistic data – relevant quality characteristics – to
form a complete basis for evaluating the product quality,
they also acknowledged the need for a methodology
change. As seen in Fig. 5, our interpretation is that for the
control system to exploit the available information fully, a
“pre-measurement” and a “post-measurement” intelligent
models should be considered for strategic inspection
protocols.
In Babu et al.’s [3] model, the usage of a connected
646 Int. J. of Automation Technology Vol.15 No.5, 2021
Multi-Layer Quality Inspection System Framework for Industry 4.0
database (for holistic data) is used as a similar advantage
that the inspector has in Juran’s model, in which the
inspector is self-aware of the production performance
and the flaws [21], to estimate deviations and, thereby,
create a measuring strategy prior to quality inspection,
i.e., pre-measurement. The measuring strategy considers
not only the on-line constraints (e.g., time available
or measurement quality) put by the designer or
engineer connected to the database but also historical
measurements or simulated CAD/CAM data to predict
part deviation patterns. Thereby, this creates a partial
measuring strategy that only focuses on what cannot be
predicted or is uncertain. For that, the quality inspection
is triggered by machine monitoring when encountering
product or process deviations as in [40].
Post-measurement implies the essential synthesis of
measurement, monitoring, and database data (e.g.,
CAD/CAM/CAE) for producing more consistent, accu-
rate, and useful information [35]. Nevertheless, all
sources of data are subject to uncertainties or assumptions
that can damage the accuracy of the conformance
(e.g., measurement procedure uncertainty). Hence, a
comparison of the results in regards to the “designer
standards” is essential. The conformance of the quality
inspection, i.e., “OK” or “NOT-OK,” should not only
consider the comparison but also calculate the level of
uncertainty in the system. In the case of “NOT-OK,” the
intelligence should decide based on the quality constraints
set by the designer if the un-conformance is due to the
quality inspection and not the product, and react on-line
with new instructions for the quality inspection system.
Finally, the control system will create the necessary
documentation of the measuring process in the database
(for future usage and self-optimisation).
4.3. Cloud-Connected Resources
For the system to be able to handle the key features
presented above, the different resources that constitute
quality inspection should be connected. AI has proven
to be a promising solution for quality inspection;
nonetheless, for the system to benefit from the analysis
and decision making of AI, high computational power
at low latency is desired. Following the new advances
in ICT, different researches have indented to integrate
connectivity (especially cloud computing) into the quality
inspection system [37, 38]. The accessibility of the
quality inspection system to all interested parties in
the product life-cycle, such as designer or production
engineers, will allow not only adjustments based on new
needs but also higher transparency [33].
5. Proposed Framework and Discussion
Above, we have discussed different research trends
and our interpretation of the relationship of holistic data,
intelligent control system, and cloud computing to the
quality inspection system. In this section, we discuss
Fig. 6. Five layers for quality inspection.
the connection between the different research trends. A
framework linking the different constituting parts together
is presented.
The usage of quality inspection techniques provides
insights from the manufacturing system by measuring one
or more product characteristics to confirm compliance
with the requirements. In the realm of technological
advances and the increase of available data, higher
demand is placed on the quality control and inspection
system to guarantee a zero defect manufacturing
system [2, 7, 31]. The quality inspection resources need
to seamlessly interact in the form of actions and data
to decide upon conformance successfully. Conformance
of the product geometrical characteristics is directly
linked with the performance of the quality inspection
system, as well as the previous value-added activities
performed prior to inspection. Holistic decision on
quality conformance considers the performance of the
activities as well as the geometrical data (features)
extracted by the measuring instrument.
Utilising well-known quality inspection
frameworks [18, 21, 25] as canvas together with the
interpretation of the identified research trends, a quality
inspection framework for Industry 4.0 is developed. In
the proposed multi-layer quality inspection framework,
see Fig. 3, the different resources, activities, and data that
compound quality inspection are performed in both cyber
and physical form; each representing a layer; see Fig. 6.
Interoperability plays a major role in the enhancement
of the quality inspection system. Cloud computing is
the promising technology described in the literature for
connecting the different layers. Nevertheless, no system
is perfect or free of errors, e.g., because of the human
intervention in the system from its design, it is susceptible
to uncertainties in the form of assumptions; see Fig. 6.
The resources layer, represented as “rectangles,”
embodies:
The work-piece to be inspected,
Measuring instrument(s),
An actuator that manipulates the measurement
instrument and possibly the work-piece,
An intelligent control system,and
A cloud-connected database.
Int. J. of Automation Technology Vol.15 No.5, 2021 647
Azamfirei, V., Granlund, A., and Lagrosen, Y.
The actions layer, represented as “black arrows,” copes
with the interaction between the different resources. The
physical actions are:
Work-piece identification,
Precise alignment of the work-piece by an internal or
external actuator or digital fixture [14, 26],
Manipulation of the measurement instrument [31],
and
Measuring.
In parallel, cyber-actions are being carried out once know-
ing the work-piece to be inspected, i.e., pre-measurement:
Estimate deviation based on specifications, monitor-
ing data, simulated data, and historical measuring
data [21, 26, 35], and
Select or generate of measuring strategy based
on the estimated deviations and given inspection
constraints [36];
and once completed the measurements using the previous
model, i.e., post measurement:
Data fusion of holistic collected data [35],
Compare to standards/specifications (geometrical
and process),
Decide upon conformance,
Process upon conformance result, if “OK,” send
to next station, if “NOT-OK,” send to repair or
re-inspect (depending on inspection constraints, e.g.,
available time, times re-inspected), and
Self-optimisation of the system, comparable to
experience and self-learning of operators [21, 30,
39].
The data layer encompasses:
Extracted data such as work-piece ID, measuring
data, and monitoring data [16, 17, 30],
Available data, such as product and process specifi-
cations, historical measuring data, and performance
of the value-added system [26, 28, 33, 36],
Insertable data on need, such as inspection
constraints (e.g., time or accuracy) [26], and
Generated data, such as discrete instructions for
measurement based on the estimated deviations [3],
the decision upon conformance, and documentation
of the measuring process and results.
It is built on the assumptions:
As Juran [21] described the inspector as part of a
socio-technological environment, and thereby, it is
aware of the value-added activities performance and
people involved prior to the quality inspection.
The quality inspection application is started with
pre-conceptions on the possible defects and what
to search for (expressed as pre-measurement
instructions).
Both data used as collected by the measurement
instrument or on-line monitoring device should be
validated (reliable), holistic, and available on-line,
i.e., real-time when needed.
Nonetheless, the entire measuring system is affected by
uncertainty, as depicted in Figs. 3 and 6. Uncertainty is
interpreted as the added distortion from reality. In quality
literature, it is expressed as to how accurate the process
(measurement, analysis, decision upon conformity...) is
from reality and how repeatable it is. The physical world
is affected by the environment and the cyber world by
the assumptions of those who have created the standards
and procedures. Real-time holistic data analysed by a
self-optimising intelligence is one of the solutions to cope
with uncertainty in deciding upon product and process
conformance that repeats in literature.
6. Conclusions and Future Work
Traditional quality methods, such as quality inspection,
suffer from significant limitations in highly customised
small batch production. For quality inspection to
remain fundamental for zero defect manufacturing and
Industry 4.0, an increase in flexibility and decision upon
conformance reliability is needed. Much research has
pointed to a methodological change need for performing
quality inspection [1]. This work presents an overview of
the latest advances in the quality inspection field in regard
to Industry 4.0. The relation of the identified research
trends is presented in a framework. From the literature
review, we have identified the need for enhancing quality
inspection with:
Holistic data (historical, captured, and generated),
Intelligent control system, and
Cloud-connected quality inspection resources.
In contrast to previous research [11–17], we have argued
that quality inspection consists of multi-layers with
respective levels of profundity. We believe it is needed to
differentiate quality inspection into five layers, in which
resources, actions, and data interact in physical and cyber
form, each representing a layer. The framework is built on
the assumption that data (used and collected) need to be
validated, holistic, and on-line, i.e., when needed, for the
system to effectively decide upon conformity to surpass
the challenges of reliability, flexibility and autonomy.
We believe that the full implementation of the proposed
quality inspection system framework enables:
Increase flexibility on inspection requirements [26,
30, 33, 36] by first, allowing the designer and pro-
duction engineer to be part of the system and import
648 Int. J. of Automation Technology Vol.15 No.5, 2021
Multi-Layer Quality Inspection System Framework for Industry 4.0
newer quality inspection constraints, standards or
instructions; second, creating a measuring strategy
based on new needs and production state; and third,
increasing communication and learning between
quality inspection resources.
Increase systems autonomy as drastic geometrical
deviations can lead to disruption of the system,
mainly due to advanced planning of inspection
procedure [12] and the system will be capable of
self-learning, self-adapting, and self-optimising [30,
39].
Decrease operation time by estimating geometrical
defects based on production state, historical mea-
surements on similar product type, or simulated
measurements [3, 17, 35, 40].
Increase visibility by constantly in real-time up-
dating the common database using cloud comput-
ing [37, 38].
Increase accuracy and reliability, i.e., decision upon
conformity based on holistic data [1, 3, 16, 28, 30]
using intelligent control system [14, 38, 39].
.
Future work will focus on implementing and validating
the proposed framework in several case studies in
different industries. This will, thereby, demonstrate
its ability to smart inspect (what and how needed),
detect false negative or positive, and improve over
time; thus, preventing defects from propagating in the
manufacturing system by taking advantage of the digital
transformation. In addition, we believe this contribution
would help with substantial advantages for introducing
quality inspection techniques not available until now or
with major drawbacks in industry, e.g., in-line quality
inspection. In order to exploit the full advantages of
the proposed framework, future research in the field of
robotic in-line quality inspection is being carried out.
Acknowledgements
This work was performed at M¨alardalen University and
alardalen Industrial Technology Center (MITC) as part of the
Industrial Research School ARRAY, financed by the Knowledge
Foundation (KKS).
References:
[1] D. Imkamp, J. Berthold, M. Heizmann, K. Kniel, M. Peterek,
R. Schmitt, J. Seidler, and K. D. Sommer, “Challenges and trends
in manufacturing measurement technology – the “Industrie 4.0”
concept,” Technisches Messen, Vol.83, No.7-8, pp. 417-429, 2016.
[2] K. S. Wang, “Towards zero-defect manufacturing (ZDM)-a data
mining approach,” Advances in Manufacturing, Vol.1, No.1,
pp. 62-74, 2013.
[3] M. Babu, P. Franciosa, and D. Ceglarek, “Spatio-Temporal
Adaptive Sampling for effective coverage measurement planning
during quality inspection of free form surfaces using robotic 3D
optical scanner,” J. of Manufacturing Systems, Vol.53, pp. 93-108,
2019.
[4] F. M. Gryna and J. M. Juran (Eds.), “Juran’s quality control
handbook,” McGraw-Hill, 4th edition, 1988.
[5] M. Uekita and Y. Takaya, “On-machine dimensional measurement
of large parts by compensating for volumetric errors of machine
tools,” Precision Engineering, Vol.43, pp. 200-210, 2016.
[6] M. Colledani, D. Coupek, A. Verl, J. Aichele, and A. Yemane,
“A cyber-physical system for quality-oriented assembly of
automotive electric motors,” CIRP J. of Manufacturing Science and
Technology, Vol.20, pp. 12-22, 2018.
[7] G. May and D. Kiritsis, “Zero Defect Manufacturing Strategies
and Platform for Smart Factories of Industry 4.0,” Proc. of the 4th
Int. Conf. on the Industry 4.0 Model for Advanced Manufacturing,
pp. 142-152, 2019.
[8] S. Phuyal, D. Bista, and R. Bista, “Challenges, Opportunities and
Future Directions of Smart Manufacturing: A State of Art Review,
Sustainable Futures, Vol.2, 100023, 2020.
[9] M. A. Akhloufi and B. Verney, “Fusion framework for 3D
inspection and thermal NDT,” SAE Technical Papers, 7, 2013.
[10] M. Papananias, T. E. McLeay, M. Mahfouf, and V. Kadirka-
manathan, “A Bayesian framework to estimate part quality and
associated uncertainties in multistage manufacturing,” Computers
in Industry, Vol.105, pp. 35-47, 2019.
[11] V. K. Pathak, A. K. Singh, M. Sivadasan, and N. K. Singh,
“Framework for Automated GD&T Inspection Using 3D Scanner,
J. of The Institution of Engineers (India): Series C, Vol.99, No.2,
pp. 197-205, 2018.
[12] O. Anokhin and R. Anderl, “Towards design for cyber-physical
inspection,” Procedia CIRP, Vol.84, pp. 400-405, 2019.
[13] P. Cicconi and R. Raffaeli, “An Industry 4.0 Framework for the
Quality Inspection in Gearboxes Production,” Proc. of CAD’19,
pp. 97-100, 2019.
[14] D. Evangelista, M. Antonelli, A. Pretto, C. Eitzinger, M. Moro,
C. Ferrari, and E. Menegatti, “SPIRIT – A Software Framework
for the Efficient Setup of Industrial Inspection Robots,” Proc. of the
2020 IEEE Int. Workshop on Metrology for Industry 4.0 and IoT,
MetroInd 4.0 and IoT 2020, pp. 622-626, 2020.
[15] J. Schmitt, J. B¨onig, T. Borggr¨afe, G. Beitinger, and J. Deuse,
“Predictive model-based quality inspection using Machine Learning
and Edge Cloud Computing,” Advanced Engineering Informatics,
Vol.45, 101101, 2020.
[16] Y. Takaya, “In-Process and On-Machine Measurement of
Machining Accuracy for Process and Product Quality Management:
A Review,” Int. J. Automation Technol., Vol.8, No.1, pp. 4-19,
2013.
[17] Y. Li, C. Liu, J. X. Gao, and W. Shen, “An integrated feature-based
dynamic control system for on-line machining, inspection and
monitoring,” Integrated Computer-Aided Engineering, Vol.22,
No.2, pp. 187-200, 2015.
[18] Chrysler Group LLC, Ford Motor Company, and General Motors
Corporation, “Measurement system analysis (MSA),” AIAG, 4th
edition, 2010.
[19] ISO, “ISO 2859-1:1999(en) Sampling procedures for inspection
by attributes – Part 1: Sampling schemes indexed by acceptance
quality limit (AQL) for lot-by-lot inspection,” Technical Report,
1999.
[20] ISO, “ISO 2859-2:2020(en) Sampling procedures for inspection by
attributes – Part 2: Sampling plans indexed by limiting quality (LQ)
for isolated lot inspection,” Technical Report, 2020.
[21] J. M. Juran, “Management of inspection and quality control,
Harper & Brothers, 1945.
[22] T. Pfeifer, “Production metrology,” Walter de Gruyter GmbH & Co
KG, 2015.
[23] C. Karlsson (Ed.), “Research Methods for Operations Manage-
ment,” Routledge, 2016.
[24] ISO, “ISO 17020 Conformity assessment – Requirements for
the operation of various types of bodies performing inspection,”
Technical Report, 2012.
[25] L. Pendrill, “Man as a Measurement Instrument,” NCSLI Measure,
Vol.9, No.4, pp. 24-35, 2014.
[26] M. Germani, F. Mandorli, M. Mengoni, and R. Raffaeli, “CAD-
based environment to bridge the gap between product design and
tolerance control,” Precision Engineering, Vol.34, No.1, pp. 7-15,
2010.
[27] A. E. Jaramillo, P. Boulanger, and F. Prieto, “On-line 3-D system for
the inspection of deformable parts,” Int. J. Adv. Manuf. Technol.,
pp. 1053-1063, 2011.
[28] J. Molleda, J. L. Car ´us Cand´as,R.Usamentiaga,D.F.Garc´ıa, J. C.
Granda, and J. L. Rendueles, “A fast and robust decision support
system for in-line quality assessment of resistance seam welds in
the steelmaking industry,” Computers in Industry, Vol.63, No.3,
pp. 222-230, 2012.
[29] K. Wang and Q. Yu, “Product quality inspection combining with
structure light system, data mining and RFID technology,” IFIP
Advances in Information and Communication Technology, Vol.411,
pp. 205-220, 2013.
Int. J. of Automation Technology Vol.15 No.5, 2021 649
Azamfirei, V., Granlund, A., and Lagrosen, Y.
[30] L. Stroppa, P. Castellini, and N. Paone, “Self-Optimizing Robot
Vision for Online Quality Control,” Experimental Techniques,
Vol.40, No.3, pp. 1051-1064, 2015.
[31] E. Kiraci, P. Franciosa, G. A. Turley, A. Olifent, A. Attridge, and
M. A. Williams, “Moving towards in-line metrology: evaluation
of a Laser Radar system for in-line dimensional inspection for
automotive assembly systems,” Int. J. of Advanced Manufacturing
Technology, Vol.91, No.1-4, pp. 69-78, 2017.
[32] R. S ¨oderberg, K. W ¨armefjord, J. S. Carlson, and L. Lindkvist,
“Toward a Digital Twin for real-time geometry assurance
in individualized production,” CIRP Annals – Manufacturing
Technology, Vol.66, No.1, pp. 137-140, 2017.
[33] V. Majstorovic, S. Stojadinovic, S. Zivkovic, D. Djurdjanovic,
Z. Jakovljevic, and N. Gligorijevic, “Cyber-Physical Manufacturing
Metrology Model (CPM3) for Sculptured Surfaces – Turbine Blade
Application,” Procedia CIRP, Vol.63, pp. 658-663, 2017.
[34] W. Gao, H. Haitjema, F. Z. Fang, R. K. Leach, C. F. Cheung,
E. Savio, and J. M. Linares, “On-machine and in-process surface
metrology for precision manufacturing,” CIRP Annals, Vol.68,
Issue 2, pp. 843-866, 2019.
[35] W. P. Syam, K. Rybalcenko, A. Gaio, J. Crabtree, and R. K.
Leach, “Methodology for the development of in-line optical surface
measuring instruments with a case study for additive surface
finishing,” Optics and Lasers in Engineering, Vol.121, pp. 271-288,
2019.
[36] N. D. M. Phan, Y. Quinsat, S. Lavernhe, and C. Lartigue, “Scanner
path planning with the control of overlap for part inspection with an
industrial robot,” Int. J. of Advanced Manufacturing Technology,
Vol.98, No.1-4, pp. 629-643, 2018.
[37] F. Xu, F. Yang, X. Wu, Q. Guo, and C. Zhao, “Application and
experiments of 5g technology powered industrial internet,” Proc.
of the 2019 IEEE Int. Conf. on Communications Workshops, ICC
Workshops, pp. 1-6, 2019.
[38] X. Zhang, Y. Cheng, L. Hu, S. Wei, and F. Wu, “Machine Vision
On-line Detection System: Applications and Standardization Re-
quirements,” Proc. of 2020 IEEE Int. Conf. on Artificial Intelligence
and Computer Applications (ICAICA 2020), pp. 1200-1204, 2020.
[39] H. Ding, R. X. Gao, A. J. Isaksson, R. G. Landers, T. Parisini, and
Y. Yuan, “State of AI-Based Monitoring in Smart Manufacturing
and Introduction to Focused Section,” IEEE/ASME Trans. on
Mechatronics, Vol.25, No.5, pp. 2143-2154, 2020.
[40] Y. Li, C. Liu, X. Hao, J. X. Gao, and P. G. Maropoulos, “Responsive
fixture design using dynamic product inspection and monitoring
technologies for the precision machining of large-scale aerospace
parts,” CIRP Annals – Manufacturing Technology, Vol.64, No.1,
pp. 173-176, 2015.
Name:
Victor Azamfirei
Affiliation:
Ph.D. Student, Division of Product Realization,
School of Innovation, Design and Technology,
alardalen University
Address:
15 Hamngatan, Eskilstuna 632 20, Sweden
Brief Biographical History:
2018 Received B.Sc. degree in Product Design Engineering from Sk ¨ovde
University
2019 Received M.Sc. degree in Industrial Systems from Sk ¨ovde
University
2019- Ph.D. Student, M ¨alardalen University
Main Works:
In-line quality inspection, robotics, assembly operations
Membership in Academic Societies:
Produktion 2030
Name:
Anna Granlund
Affiliation:
Senior Lecturer, Division of Product Realization,
School of Innovation, Design and Technology,
alardalen University
Address:
15 Hamngatan, Eskilstuna 632 20, Sweden
Brief Biographical History:
2014 Received Ph.D. in Innovation and Design from M¨alardalen
University
2014-2018 Researcher, School of Innovation, Design and Technology,
alardalen University
2018- Senior Lecturer, School of Innovation, Design and Technology,
alardalen University
Main Works:
A. Granlund, C. R ¨osi ¨o, J. Bruch, and P. E. Johansson, “Lead factory
operationalisation and challenges,” Production planning and control,
Vol.30, No.2-3, pp. 96-111, 2019.
A. Granlund and M. Wiktorsson, “Automation in internal logistics:
strategic and operational challenges,” Int. J. of Logistics Systems and
Management, Vol.18, No.4, pp. 538-558, 2014.
Membership in Academic Societies:
European Operations Management Association (EurOMA)
Design Society (DS)
Name:
Yvonne Lagrosen
Affiliation:
Associate Professor, Division of Product Re-
alization, School of Innovation, Design and
Technology, M ¨alardalen University
Address:
15 Hamngatan, Eskilstuna 632 20, Sweden
Brief Biographical History:
2012- Associate Professor of Quality Management
Main Works:
Y. Lagrosen and S. Lagrosen, “Creating a culture for sustainability and
quality – a lean-inspired way of working,” Total Quality Management &
Business Excellence, pp. 1-15, doi: 10.1080/14783363.2019.1575199,
2019.
Y. Lagrosen, F. Travis, and S. Lagrosen, “Brain integration as a driver for
quality management success,” Int. J. of Quality and Service Sciences,
Vol.4, No.3, pp. 253-269, 2012.
Membership in Academic Societies:
Swedish Quality Management Association (SQMA)
650 Int. J. of Automation Technology Vol.15 No.5, 2021
Powered by TCPDF (www.tcpdf.org)
... This process entails extensive manual checking, which is not only laborious and time-intensive but also susceptible to errors. Therefore, the inspection of critical steps is highly important during the assembly process, while the inspection of assembly quality in manual assembly has become a bottleneck, limiting both product quality improvement and the enhancement of batch production capacity [6]. The rapid advancement of machine vision technologies utilizing deep learning methodologies has made the benefits of machine vision-based inspection increasingly apparent, such as high accuracy and precision, improving inspection efficiency, reducing the need for manual labor, and cutting costs. ...
... 5 I can perform the tasks by the status information provided by this condition. 6 This interface brings me a sense of amazement. 7 I can receive timely notifications during inspecting tasks. ...
Article
Full-text available
In the realm of industrial quality control, visual inspection plays a pivotal role in ensuring product precision and consistency. Moreover, it enables non-contact inspection, preventing the products from potential damage, and timely monitoring capabilities facilitate quick decision making. However, traditional methods, such as manual inspection using feeler gauges, are time-consuming, labor-intensive, and prone to human error. To address these limitations, this study proposes a deep learning-based visual inspection system for measuring gap spacing in high-precision equipment. Utilizing the DeepLSD algorithm, the system integrates traditional and deep learning techniques to enhance line segment detection, resulting in more robust and accurate inspection outcomes. Key performance improvements were realized, with the proposed system being a piece of deep learning-enabled high-precision mobile equipment for inspecting gap spacing in real-time. Through a comparative analysis with the traditional feeler gauge method, the proposed system demonstrated significant improvements in inspection time, accuracy, and user experience, while reducing workload. Experimental results validate the effectiveness and efficiency of the proposed approach, highlighting its potential for widespread application in industrial quality inspection activities.
... Recently, many papers about ZDM procedures have been published in different fields, ranging from general papers [1,11,12] to papers focusing on quality inspection [24]. This survey can help companies determine which processes need to be developed for successful implementation of ZDM [1,11,12] [25] The readiness level of a company's culture for adopting ZDM is reflected in the consciousness of its leaders. The use of yes/no questionnaires in the proposed tool objectively shows the true level of cultural readiness [18,25], and the amount of investment needed for implementing ZDM varies based on the level of readiness. ...
... This survey can help companies determine which processes need to be developed for successful implementation of ZDM [1,11,12] [25] The readiness level of a company's culture for adopting ZDM is reflected in the consciousness of its leaders. The use of yes/no questionnaires in the proposed tool objectively shows the true level of cultural readiness [18,25], and the amount of investment needed for implementing ZDM varies based on the level of readiness. To avoid subjectivity an expert outside the company should conduct the questionnaire [19]. ...
Article
Full-text available
In this study, a comprehensive framework for assessing the readiness of production systems for Zero Defect Manufacturing (ZDM) has been developed and presented. The framework includes four pillars of ZDM readiness, namely Personnel, Procedures, Infrastructure, and Company Culture, to help companies understand their level of readiness and plan for successful implementation of ZDM. We argue that a manufacturing company will be better equipped to embrace ZDM if it performs well in these four areas. We propose a tool that uses yes/no questionnaires to assess a manufacturing system's readiness for ZDM. The results of the questionnaire will objectively show the true level of cultural readiness for ZDM adoption, and the level of investment required for implementation will depend on the level of readiness. This tool can help companies gain a clear understanding of their readiness and create a plan for implementing ZDM. Overall , our framework and tool can help manufacturers improve the quality of their products and be ready for ZDM adoption.
... In the context of intelligent manufacturing, researching an advanced inspection strategy and solution to replace the traditional low-efficiency inspection model has become one of the keys for steel companies to improve quality and efficiency. Advanced quality inspection strategies and approaches not only reduce inspection costs but also bring overall manufacturing cost advantages [5,6]. ...
Article
Full-text available
Pre-production quality defect inspection is a crucial step in industrial manufacturing, and many traditional inspection strategies suffer from inefficiency issues. This is especially true for tasks such as mechanical performance testing of steel products, which involve time-consuming processes like offline sampling, specimen preparation, and testing. The inspection volume significantly impacts the production cycle, inventory, yield, and labor costs. Constructing a data-driven model for predicting product quality and implementing proactive sampling inspection based on the prediction results is an appealing solution. However, the prediction uncertainty of data-driven models poses a challenging problem that needs to be addressed. This paper proposes an active quality inspection approach for steel products based on the uncertainty quantification in the predictive model for mechanical performance. The objective is to reduce both the sampling frequency and the omission rate on the production site. First, an ensemble model based on improved lower and upper bound estimation is established for interval prediction of mechanical performance. The uncertainty of the specific value prediction model is quantitatively estimated using interval probability distributions. Then, a predictive model for the mechanical performance failure probability is built based on the prediction interval size and probability distribution. By determining an appropriate probability threshold, the trade-off between prediction accuracy and defect detection accuracy (recall rate) is balanced, enabling the establishment of an active sampling strategy. Finally, this functionality is integrated into the manufacturing execution system of a steel factory, realizing a mechanical performance inspection approach based on proactive sampling. The proposed approach is validated using real production datasets. When the probability threshold is set to 30%, the prediction accuracy and recall rate for failure mechanical performance samples are 75% and 100%, respectively. Meanwhile, the sampling rate is only 5.33%, while controlling the risk of omission. This represents a 50% reduction in sampling rate compared to the inspection rules commonly used in actual production. The overall efficiency of product quality inspection is improved, and inspection costs are reduced.
... Psarommatis et al. [6] reviewed the state-of-the-art of ZDM and highlighted the importance of shifting from local to global solutions while considering the role of people and human activities in manufacturing. Azamfirei et al. [12] reviewed quality inspection protocols and proposed a multi-layer framework for ZDM, characterised by holistic data, intelligent control system, and connectivity. Babic et al. [13] reviewed the usage of image-based quality inspection in smart manufacturing. ...
...  Defect identification and sorting: Robots equipped with machine vision and robotic arms can identify defective parts and sort them accordingly. Once defects are detected during the quality-inspection process, the robots can remove or separate the faulty parts from the production line, ensuring that only high-quality components are delivered [145].  Data analysis and feedback loop: Automation systems can collect and analyse vast amounts of data generated during CNC machining operations. ...
Article
Full-text available
Robotics and automation have significantly transformed CNC machining operations, enhancing productivity, precision, and efficiency. Robots are employed to load and unload raw materials, workpieces, and finished parts onto CNC machines. They can efficiently handle heavy and bulky components, reducing the demand of manual labor and minimizing the risk of injuries. Robots can also be used in CNC machine tools to perform tasks such as automatic tool changing system, part inspection, and workpiece positioning. Automation technologies, including in-line inspection systems and non-destructive testing methods, can be integrated into CNC machining cells to enhance accuracy and reduce scrap and rework in machining operations. These systems collect real-time data on process parameters and machine tool performance to predict maintenance, optimize machining parameters, and improve overall efficiency. In the study, applications of robotics and automation in modification of CNC machine tools are reviewed and discussed. Different applications of robotics and automations in CNC machine tools such as automated material handling, automatic tool changing, robotic work cells, adaptive machining, machine tending, quality inspection, data monitoring and analysis and production line integration are discussed. Thus, by analyzing recent achievements in published papers, new ideas and concepts of future research works are suggested. As a result, accuracy as well as productivity in process of part production can be enhanced by applying the robotics and automation in CNC machining operations.
Article
With the Fourth Industrial Revolution (Industry 4.0) and advances in digital technology, zero-defect manufacturing (ZDM) has become a transformative and attractive concept that has the potential to reshape the manufacturing landscape. In this structured literature review, recent developments in ZDM in the textile industry from 2004 to 2023 are examined, with a focus on detection, repair, prediction, and prevention. Through bibliometrics analysis and evaluation of the current situation of ZDM technology, four main shortcomings are highlighted, to be specific, limitations in automated defect detection, incomplete artificial intelligence (AI)-based repair strategies, restricted predictive research, and focused prevention mechanism. Meanwhile, open challenges that require urgent attention are explored, that is systematic ZDM strategy integration, data management complexity, and demand for flexible ZDM frameworks. To address these shortcomings and challenges, three further prospectives are proposed, including addressing research imbalance, vision for an integrated ZDM system, and evolutionary predictive models. These prospectives aim to advance the field and drive the holistic development of ZDM technologies in the textile industry by promoting a more intelligent production strategy with higher quality and less waste.
Article
Full-text available
The supply of defect-free, high-quality products is an important success factor for the long-term competitiveness of manufacturing companies. Despite the increasing challenges of rising product variety and complexity and the necessity of economic manufacturing, a comprehensive and reliable quality inspection is often indispensable. In consequence, high inspection volumes turn inspection processes into manufacturing bottlenecks. In this contribution, we investigate a new integrated solution of predictive model-based quality inspection in industrial manufacturing by utilizing Machine Learning techniques and Edge Cloud Computing technology. In contrast to state-of-the-art contributions, we propose a holistic approach comprising the target-oriented data acquisition and processing, modelling and model deployment as well as the technological implementation in the existing IT plant infrastructure. A real industrial use case in SMT manufacturing is presented to underline the procedure and benefits of the proposed method. The results show that by employing the proposed method, inspection volumes can be reduced significantly and thus economic advantages can be generated.
Article
Full-text available
Smart manufacturing is the technology utilizing the interconnected machines and tools for improving manufacturing performance and optimizing the energy and workforce required by the implementation of big data processing, artificial intelligence and advanced robotics technology and interconnectivity of them. This paper defines and discusses the smart manufacturing system and states it current implementation status and analyzes the gap between current manufacturing system and the predicted future smart manufacturing system, discusses the technologies associated with it and their contribution in smart manufacturing technology. Also, to realize this rapidly growing technology and cover its all dimensions a survey of the latest developments in this field and its impacts were analyzed and presented along with the challenges of implementation, opportunities and the future directions for smart manufacturing system.
Article
Full-text available
Our latest research at the Digital Lifecycle Management group, WMG, makes possible effective in-line application of robotic 3D-optical scanners for quality inspection of large free form surfaces. The Spatio-Temporal Adaptive Sampling (STAS) methodology developed in this research overcomes the limitation of large measurement cycle time of 3D-surfaces scanners and has the following key capabilities: 1. Can accurately predict whole part deviations using partial measurements. 2. Enables intelligent automation of the 3D-scanner by identifying critical regions to measure. 3. Enables real-time product quality monitoring in a smart manufacturing environment. 4. Achieves a cycle time reduction of 42.18% compared to measuring the entire part.
Article
Full-text available
The ongoing technological advances in information and communication technology have a significant impact on the field of product creation. Their implementation in manufacturing equipment yields cyber-physical production systems, which promise significant increase in flexibility and efficiency concerning value chains and complex manufacturing networks. To ensure the high quality of the designed and manufactured products new ways of inspection is required. This work discusses the use of computer-aided models and tools to realize a cyber-physical inspection. Therefore, a two-staged methodology was applied. At first a brief overview over basic overall approaches and current work in the field of inspection processes as well as promising advances in the field of design and information exchange are presented. The second step analyses the Siemens PLM NX 12 software and correlates the provided functionality with the necessary activities defined in the first phase. Following, the paper discusses arising challenges and necessary extensions to move towards a cyber-physical inspection. Finally a brief conclusion as well as a description of future work is given.
Article
Over the past few decades, intelligentization, supported by artificial intelligence (AI) technologies, has become an important trend for industrial manufacturing, accelerating the development of smart manufacturing. In modern industries, standard AI has been endowed with additional attributes, yielding the so-called industrial artificial intelligence (IAI) that has become the technical core of smart manufacturing. AI-powered manufacturing brings remarkable improvements in many aspects of closed-loop production chains from manufacturing processes to end product logistics. In particular, IAI incorporating domain knowledge has benefited the area of production monitoring considerably. Advanced AI methods such as deep neural networks, adversarial training, and transfer learning have been widely used to support both diagnostics and predictive maintenance of the entire production process. It is generally believed that IAI is the critical technologies needed to drive the future evolution of industrial manufacturing. This article offers a comprehensive overview of AI-powered manufacturing and its applications in monitoring. More specifically, it summarizes the key technologies of IAI and discusses their typical application scenarios with respect to three major aspects of production monitoring: fault diagnosis, remaining useful life prediction, and quality inspection. In addition, the existing problems and future research directions of IAI are also discussed. This article further introduces the papers in this focused section on AI-based monitoring in smart manufacturing by weaving them into the overview, highlighting how they contribute to and extend the body of literature in this area.
Article
On-machine and in-process surface metrology are important for quality control in manufacturing of precision surfaces. The definitions, requirements and tasks of on-machine and in-process surface metrology are addressed. The state-of-the-art on-machine and in-process measurement systems and sensor technologies are presented. Error separation algorithms for removing machine tool errors, which is specially required in on-machine and in-process surface metrology, are overviewed, followed by a discussion on calibration and traceability. Advanced techniques on sampling strategies, measurement systems-machine tools interface, data flow and analysis as well as feedbacks for compensation manufacturing are then demonstrated. Future challenges and developing trends are also discussed.