Conference PaperPDF Available

Recent Advances and Trends of Cyber-Physical Systems and Big Data Analytics in Industrial Informatics

Authors:

Abstract and Figures

In today's competitive business environment, companies are facing challenges in dealing with big data issues for rapid decision making for improved productivity. Many manufacturing systems are not ready to manage big data due to the lack of smart analytics tools. Germany is leading a transformation toward 4th Generation Industrial Revolution (Industry 4.0) based on Cyber-Physical System based manufacturing and service innovation. As more software and embedded intelligence are integrated in industrial products and systems, predictive technologies can further intertwine intelligent algorithms with electronics and tether-free intelligence to predict product performance degradation and autonomously manage and optimize product service needs. This article addresses the trends of industrial transformation in big data environment as well as the readiness of smart predictive informatics tools to manage big data to achieve transparency and productivity.
Content may be subject to copyright.
Proceeding of Int. Conference on Industrial Informatics (INDIN) 2014
Recent Advances and Trends of Cyber-Physical
Systems and Big Data Analytics in Industrial
Informatics
Jay Lee, Behrad Bagheri, Hung-An Kao
Center for Intelligent Maintenance Systems
University of Cincinnati
Cincinnati, United States
Abstract In today’s competitive business environment,
companies are facing challenges in dealing with big data issues
for rapid decision making for improved productivity. Many
manufacturing systems are not ready to manage big data due to
the lack of smart analytics tools. Germany is leading a
transformation toward 4th Generation Industrial Revolution
(Industry 4.0) based on Cyber-Physical System based
manufacturing and service innovation. As more software and
embedded intelligence are integrated in industrial products and
systems, predictive technologies can further intertwine intelligent
algorithms with electronics and tether-free intelligence to predict
product performance degradation and autonomously manage
and optimize product service needs. This article addresses the
trends of industrial transformation in big data environment as
well as the readiness of smart predictive informatics tools to
manage big data to achieve transparency and productivity.
KeywordsIndustry 4.0; Cyber Physical Systems; Prognostics
and Health Management; Big Data;
I. INTRODUCTION
In the past decade, emergence of promising tools such as
Enterprise Systems provided companies with solutions to
improve their productivity and service quality. But today's
competitive nature of world industry enforces companies to
implement more recent technologies to secure their position
among competitors [1]. Recent technological advances in the
field of communication and computer science have provided
cost-effective solutions for companies to acquire and transfer
gigantic amount of data from their fleet of assets. Consequently
Handling these huge sets of data is not easily achievable,
therefore supporting "Big Data" is one of the most recent topics
in the world industry. Introduction of methods and terms such
as internet of things and interconnected systems are among the
efforts of researchers and industrial companies to address
applicable solutions in the "Big Data" environment. Therefore,
the requirement of systematic approaches to handle and
analyze enterprise data in the Big Data environment is the
purpose of several research studies including current paper. [2].
II. RECENT ADVANCES AND TERMINOLOGIES
In such competitive and creative environment, new terms
and phrases have born to address the current requirement and
demands of the industry. Recently, Germany has announced
the Industry 4.0 methodology as the fourth industrial
revolution. Three past revolutions are "First mechanical loom"
in 1784, "First assembly line" in 1870, "First programmable
logic controller (PLC)" in 1969 respectively. Based on the
industry 4.0 terminology, intelligent data analysis and
interconnected systems are combined together to generate a
brand new aspect in factory transformation and production
management. Table 1 represents the differences between a
today's factory and an Industry 4.0 factory. In current industry
environment, providing high-end quality service or product
with the least cost is the key to success and industrial factories
are trying to achieve as much performance as possible to
increase their profit as well as their reputation. In this way,
various data sources are available to provide worthwhile
information about different aspects of the factory. In this stage,
the utilization of data for understanding the current condition
and detect faults and failures is an important topic to research.
e. g. in production, there are various commercial tools
available to provide OEE (Overall Equipment Effectiveness)
information to factory management in order to highlight root
cause of problems and possible faults in the system. In
contrast, in an Industry 4.0 factory, in addition to condition
monitoring and fault diagnosis, components and systems are
able to gain self-awareness and self-predictiveness, which will
provide management with more insight on the status of the
factory. Furthermore, peer-to-peer comparison and fusion of
health information from various components provides a precise
health prediction in component and system levels and enforce
factory management to trigger required maintenance at the best
possible time to reach just-in time maintenance and gain near
zero downtime. In addition to Industry 4.0, Cyber-physical
system (CPS) is a phrase representing the integrated
computational and physical capabilities such as sensing,
communication and actuation to physical world[3], which is
addressed by American government since 2007 as a new
developments strategy[4],[5].
Proceeding of Int. Conference on Industrial Informatics (INDIN) 2014
Applications of CPS include but not limited to
manufacturing, Secure Control, medical devices,
Environmental Control, aviation, advanced automotive
systems, process control, energy control, traffic control and
safety, smart structures and so on[6]. Having in mind that CPS
is in its immature stage and it covers a broad range of scientific
era, interactive collaborations between industry and academia,
especially long-term collaborations such as the tasks that have
been done in Advanced Research and Technology for
Embedded Intelligence Systems (ARTEMIS) in Europe or
Center for Intelligent Maintenance Systems (IMS) in United
States will help in identification of challenges and significantly
accelerate the progress of CPS[4], [7].
In the implementation of advanced terms such as Industry
4.0 and Cyber Physical Systems, sole presence of connectivity
between machines and using sensors is not useful. To leverage
these advanced technologies, correct information has to be
present at the right time for the right purpose. In this situation,
a 6C system that is consist of Connection (sensor and
networks), Cloud (data on demand), Cyber (model & memory),
Content (meaning and correlation), Community (sharing &
collaboration), and Customization (personalization and value)
[7] can enhance the information system. In this scenario and in
order to provide useful insight to the factory management and
gain correct content, data has to be processed with advanced
tools to generate relative information. Considering the presence
of visible and invisible issues in an industrial factory, the
information generation algorithm has to capable of detecting
and addressing invisible issues such as machine degradation,
component wear, etc in the factory floor.
III. PROGNOSTICS AND HEALTH MANAGEMENT
Prognostics and health management (PHM) is an enrich
research area that mostly deals with component wear and
degradation. Factory wide transparency is one of the most
important targets of PHM because in contrast with visible
issues, invisible issues might happen due machine degradation,
component wear and etc. while operators and factory managers
are not aware of them. Indentifying those issues can lead to
serious downtimes in the factory floor and PHM as an ever-
growing pioneer research domain provides useful approaches
to detect those invisible problems and avoid unplanned
downtimes to overwhelm uncertainties in the systems.
Remaining useful life prediction, fault diagnosis and fault
detection are among the useful tools provided by PHM
algorithms for bringing transparency in every aspect of future
industry by analyzing sensory and system level data. In Past
few years, different aspects of industry as well as different
machines or components have been targeted by researchers for
developing reliable PHM solutions. Aircraft engines [8],
industrial robots [9], machine tools [10], electrical motors[11],
wind turbine[12], batteries [13], gearboxes [14], bearings [15],
pumps [16] and etc are only few examples of a broad range of
assets.
IV. INTEGRATED SYSTEMS
In previous sections, we introduced Industry 4.0 and Cyber-
Physical Systems as two leading infrastructures for managing
data and leaning toward more efficient production in current
industry. In addition, we presented prognostics and health
management methods as powerful and applicable approaches
for detecting invisible faults and failures among factories fleet.
In current section, advantages of implementing PHM
algorithms in Industry 4.0 and CPS application will be
presented and the next section introduces a methodology for
implementing cyber-physical systems in industry 4.0
applications.
Today, most PHM methods have only access to target asset
data. Asset data might contain hundreds of sensory and system
data readings and provide good insight and prognostics results
for the current asset. But, there are much more sources of
information that are not considered in today's implementation
of PHM methods. Peer-to-peer evaluation through fleet of
assets, life-cycle historical data from identical assets and
system configuration are few examples of the huge portion of
data that is ignored by now.. For example, in PHM, various
failure signatures are required for accurate fault diagnosis of
assets and improve the reliability of maintenance scheme.
Table 1: Comparison of today’s factory and an Industry 4.0 factory
Proceeding of Int. Conference on Industrial Informatics (INDIN) 2014
Acquiring every possible failure signature from one instance of
assets is not likely to happen but in an interconnected system
which manages hundreds of similar assets among the fleet, the
occurrence likelihood of various failures much higher.
Therefore, the PHM analytical engine can capture those failure
signatures from various assets to refine its failure detection and
estimation capabilities. This huge improvement is possible by
utilizing the concept of cyber-physical systems for PHM
applications in which a cyber twin model (avatar) of real
machine is created and operated in cloud
platform parallel to actual asset. Leveraging the Integrated
knowledge available in cyber physical model, health condition
of assets can be accurately simulated and appropriately
presented to dedicated users upon their demand without
geographical limitations. Furthermore, the interconnectivity of
fleet of identical assets to their cyber-physical model not only
provides the opportunity of peer-to-peer evaluation and
prognostic library accumulation but also enables PHM
algorithms to have access to various life stages of different
assets as well as test stage data.
V. METHODOLOGY FOR DESIGNING CPS BASED
INDUSTRY 4.0 SYSTEMS
Knowing the capabilities of cyber-physical systems, a
promising methodology for designing Industry 4.0 applications
based on Cyber-Physical Systems can be developed. As it was
discussed in previous section, interconnectivity provides access
to vast amount of data. But, sole availability of data does not
create a significant advantage. Therefore, an adaptive yet
powerful methodology is required to manage, categorize and
process data for further analysis by PHM algorithms. This
method has to be broad enough to truly leverage all the
advantages of cyber-physical systems.
In this article, we propose “Time Machine Methodology for
Cyber-Physical Systems” which is in charge of perfectly
organize available data in Big Data environment to be prepared
for usage in PHM algorithms. Every single component of the
fleet will have a representative Time Machine record in the
cyber space. This cyber delegate extracts worthwhile
information from the pool of available data and normalizes it
for further analysis. Extracted information include but not
limited to implementation history, stress and load, operation
parameters, system configurations and maintenance records.
Once the actual component is failed, it will be removed from
the parent machine and will no longer exist to participate in
analytics. But its cyber twin (Time machine record) will
remain without any time constraints. Unlimited existence of
cyber twins results in continuous accumulation of Time
Machine records and consequently gathering various operation
parameters from broad range of identical components.
Normalization of parameters, in which further research efforts
have to be conducted, ensures the comparability of time
machine records with each other for identical components.
Additionally, time machine records obey the hierarchical
relation of actual components as well and every cyber twin has
access to records of its predecessor and ancestor components.
Such information rich environment brings significant
robustness to PHM algorithms for continuous and accurate
predicting and monitoring of the factory. Ultimately, this
methodology brings ultimate implementation of cyber physical
system into action for designing an Industry 4.0 factory.
Fig. 1. The general framework of the application of PHM algorithms in detecting invisible issues in industry using Watchdog Agent® tool
Proceeding of Int. Conference on Industrial Informatics (INDIN) 2014
VI. CASE STUDIES
A. Industrial Robot
In this section we consider the development of a cyber
physical system for health monitoring of industrial robots. Fig.
2. shows the schematic view of the cyber-physical model for
current case study on industrial robots. This study was focused
on preparing a predictive health monitoring solution for a fleet
of 30 industrial robots in production line. Due to variety of line
speeds, a more complicated multi-regime approach had to be
undertaken for establishing a robust prognostics and health
management algorithm as the core analytics of cyber physical
system by using torque and speed data. Due to its non-invasive
nature torque monitoring is a popular fault detection method
for monitoring health condition of industrial robots and
therefore, most of the research efforts in this area have been
focused on this parameter. In addition, nonlinear relation
between operating speed and torque cast a challenge on PHM
algorithms to correctly determine the health state of the robot.
In addition to condition data (torque and speed) the cyber
physical model obtains various configuration parameters such
as gear ratio, load ratio, pressure calibration, type of tooling for
robot servo guns and assigned products to specific robots from
production line.
Fig. 2. Cyber-Physical Model for Industrial Robot Monitoring
Fig. 3. Methodology for Designing CPS based Industry 4.0 Systems
Proceeding of Int. Conference on Industrial Informatics (INDIN) 2014
These configuration parameters help the core of the model to
standardize and adaptively cluster the operation data for more
accurate processing. Ultimately, the entire analytical engine of
this study was set and established on the cloud where the whole
dataset (condition and configuration data) were stored on the
cloud storage and health monitoring algorithm used the data
stored on the cloud (from entire fleet) to calculate the health
condition of each individual instance. The outcome of PHM
algorithms were presented to users as enriched infographics
through a web-based user interface
B. Virtual Battery
The battery pack is one of the most significant components
of an Electric or Hybrid Vehicle. The uncertainty in the driving
range, the batteries reliability and service life, and battery
safety concerns, are all challenges that must be overcome in
order to achieve more widespread adoption of electric vehicles.
In order to meet the challenges and rising demand in hybrid
and electric vehicles in recent years, the importance of batteries
has become more and more important[17]. Therefore, fully
understanding of the dynamic performance of batteries under
various conditions is of significant importance. However,
batteries operation condition includes several stress factors
such as environment temperature, humidity, driving style,
charging level, discharge rate and road conditions. To achieve
this battery understanding a battery model is needed which
should provide capabilities for health evaluation and failure
prediction, through simulation of cell performance under
different conditions. Additionally, using such a battery model,
the functionality of a manufactured battery can be fed back to
design and manufacturing suppliers to recognize the potential
impact of design and enhancement of manufacturing process
on battery performance[18].
An overview of the electric vehicle and battery health
management and prognostic platform is illustrated in fig. 4,
which includes algorithms for state of charge and state of
health estimation and driving behavior classification.
Presently, most of the battery models focus on individual
battery cells, or on battery package not including the detailed
dynamics within the battery. However, all components in a
battery are associated with each other. The interactions among
all the cells, conductors, BMS, and environment temperature
will play a significant role in battery performance, and deserve
an integrated study. Moreover, manufacturing practice strongly
impact on battery application conditions & user behavior.
Therefore a simulated framework is needed which is
capable of execution the functions discussed above by
integrating various models to emulate the multi-regime
changes in battery parameters and inputs and investigate their
impacts on battery functionality .
A "Virtual Battery" is a cyber-physical model resulting
from merging the new battery technology with intelligent
enabling tools to provide following solutions:
1. Enabling prognostics tools to transform data to health
information regarding the health, reliability and operational
readiness of the battery system.
2. Adequate visualization system to deliver the right
information to the right people. Some information that requires
the immediate intervention of the driver may be displayed for
the actual user (driver) on the dashboard, while other forms of
information (including more detailed diagnostic information)
would be necessary for a maintenance personnel or logistics
center and used to schedule maintenance or replacements.
3. Enabling tether-free communication to capture this
information in real-time from the vehicle and deliver it to a
central logistics center that will advise the driver about his
battery condition and could use GPS trip information to direct
the driver to the nearest service center or rapid charging center
on his trip.
4. Simulating and predicting battery behavior such as
SoC (State of Charge) and SoH (State of Health) in other
operation conditions based on external parameters(
environment , driving style, etc ) and internal parameters (
battery type, and battery age) without take a long time to test a
battery in that condition.
5. Help the designers to find the flaws or design issues
for a better hybrid cycle management in parallel with a better
estimation of the selected pack’s lifetime expectancy in the
specific application.
6. Help the battery manufacturers to locate the faulty
packs and study them to refine their design, manufacturing
process or raw material selection.
7. Integration of the wireless solutions into the
prognostic tool could also open an opportunity for better
monitoring of the vehicle during standard or extended warranty
period. The faulty system could be recalled for repair long
before the usual inspection periods. This in turn can reduce the
costs of warranty that could rise from such occasions. There is
also a possibility to integrate this system into engine
management sensors. This will enable the designers to study
the battery aging process in different climatic conditions that
the vehicle operates in. This over-time data acquisition could
be an easy and cheap way to collect field data. These data
could be used to design the next generation systems in a better
way.
VII. CONCLUSION
In this article we introduced recent advances in industrial
informatics with respect to Big Data environment, Cyber-
Physical Systems and Industry 4.0. In addition, the importance
of intelligent prognostics and health management in industry
for retaining production and service excellence has been
discussed. Furthermore, the fusion of cyber-physical systems
and PHM algorithms in big data environment and the
advantage of using interconnected systems has been presented
along with two industrial application of cyber-physical
systems.
Current industrial evolution is guiding industry toward
maximum leverage from benefits of interconnected systems in
big data environment where companies with more futuristic
vision that establish new methodologies in their culture will
have the opportunity of being significantly successful and
profitable in recent future.
Proceeding of Int. Conference on Industrial Informatics (INDIN) 2014
REFERENCES
[1] L. Da Xu, “Enterprise systems: state-of-the-art and future trends,”
Industrial Informatics, 2011.
[2] E. Dumbill, “Big Data is Rocket Fuel,Big Data, 2013.
[3] “Adaptive calibration for fusion-based cyber-physical systems,” vol. 11,
no. 4, pp. 125, Dec. 2012.
[4] “Cyber-physical systems,” IEEE Control Systems Society, 2011.
[5] J. Shi, J. Wan, H. Yan, and H. Suo, “A survey of Cyber-Physical
Systems,” Signal Processing (WCSP 2011), pp. 16, 2011.
[6] “Cyber Physical Systems: the need for new models and design
paradigms,” Carnegie Mellon University.
[7] J. Lee, E. Lapira, B. Bagheri, and H. Kao, “Recent advances and trends
in predictive manufacturing systems in big data environment,” Manufacturing
Letters, 2013.
[8] T. Wang, J. Yu, D. Siegel, and J. Lee, “A similarity-based prognostics
approach for Remaining Useful Life estimation of engineered systems,” 2008
International Conference on Prognostics and Health Management (PHM), pp.
1–6, 2008.
[9] E. R. Lapira, “Fault detection in a network of similar machines using
clustering approach,” 2012.
[10] L. Liao and J. Lee, “Design of a reconfigurable prognostics platform for
machine tools,” Expert systems with applications, vol. 37, no. 1, pp. 240252,
2010.
[11] B. Bagheri, H. Ahmadi, and R. Labbafi, “Application of data mining and
feature extraction on intelligent fault diagnosis by Artificial Neural Network
and k-nearest neighbor,” 2010 XIX International Conference on Electrical
Machines (ICEM), pp. 17, 2010.
[12] E. Lapira, D. Brisset, H. D. Ardakani, and D. Siegel, “Wind turbine
performance assessment using multi-regime modeling approach,” Renewable
Energy, 2012.
[13] M. AbuAli, E. R. Lapira, W. Zhao, D. Siegel, M. Rezvani, and J. Lee,
“Systematic Design of Prognostics and Health Management Solutions for
Energy Applications,” in Engineering Asset Management 2011, no. 22, J. Lee,
J. Ni, J. Sarangapani, and J. Mathew, Eds. London: Springer London, 2013,
pp. 243250.
[14] B. Bagheri, H. Ahmadi, and R. Labbafi, “Implementing discrete wavelet
transform and artificial neural networks for acoustic condition monitoring of
gearbox,” Elixir Mech Engg, 2011.
[15] O. D. Snell and I. Nairne, “Acoustic bearing monitoringthe future
RCM 2008,” Railway Condition Monitoring, 2008.
[16] A. Soylemezoglu and S. Jagannathan, “Mahalanobis-Taguchi System as
a Multi-Sensor Based Decision Making Prognostics Tool for Centrifugal
Pump Failures,” Reliability, IEEE Transactions on, vol.60, no.4, pp.864,878,
Dec. 2011.
[17] F. Ju, J. Wang, J. Li, and G. Xiao, “Virtual battery: A battery simulation
framework for electric vehicles,” Automation Science and Engineering, IEEE
Transactions on , vol.10, no.1, pp.5,15, Jan. 2013.
[18] Y. He, W. Liu, and B. J. Koch, “Battery algorithm verification and
development using hardware-in-the-loop testing,” Journal of Power Sources,
2010.
Fig. 4. Cyber Physical Battery Platform
... In the era of IT, many tools and machines are used for monitoring and controlling these sources. Examples on these IT tools include cyber-physical systems (Lee et al., 2014), distribution management system, geographic information systems, outage management systems, customer information systems, and Supervisory Control and Data Acquisition system (SCADA) (Al-Ali and Aburukba, 2015). ...
Article
Full-text available
The application of big data in the energy sector is considered as one of the main elements of Energy Internet. Crucial and promising challenges exist especially with the integration of renewable energy sources and smart grids. The ability to collect data and to properly use it for better decision-making is a key feature; in this work, the benefits and challenges of implementing big data analytics for renewable energy power stations are addressed. A framework was developed for the potential implementation of big data analytics for smart grids and renewable energy power utilities. A five-step approach is proposed for predicting the smart grid stability by using five different machine learning methods. Data from a decentralized smart grid data system consisting of 60,000 instances and 12 attributes was used to predict the stability of the system through three different machine learning methods. The results of fitting the penalized linear regression model show an accuracy of 96% for the model implemented using 70% of the data as a training set. Using the random forest tree model has shown 84% accuracy, and the decision tree model has shown 78% accuracy. Both the convolutional neural network model and the gradient boosted decision tree model yielded 87% for the classification model. The main limitation of this work is that the amount of data available in the data set is considered relatively small for big data analytics; however the cloud computing and real-time event analysis provided was suitable for big data analytics framework. Future research should include bigger datasets with variety of renewable energy sources and demand across more countries.
... In the era of IT, many tools and machines are used for monitoring and controlling these sources. Examples on these IT tools include cyber-physical systems (Lee et al., 2014), distribution management system, geographic information systems, outage management systems, customer information systems, and Supervisory Control and Data Acquisition system (SCADA) (Al-Ali and Aburukba, 2015). ...
Article
Full-text available
The challenge of Big data is fundamentally concerned with performing data analytics for large amount of heterogeneous data. This data can be collected from different and/or uncorrelated sources. Due to the complexity of such technology; there are still various possible applications and integrations under study particularly in the fields of smart systems with using trending technologies such as Internet of Things and Cloud computing and utilizing relevant tools and equipment such as advanced sensors and smart meters. The application of big data in the energy sector is considered as one of the main elements of Energy Internet. Crucial and promising challenges exist especially with the integration of renewable energy sources and smart grids. The ability to collect data and to properly use it for better decision-making is a key feature of smart grids. For this purpose, the convenient storing, processing, provision and analysis of information on the renewable energy system behavior is addressed. In this work, the benefits and challenges of implementing big data analytics (BDA) for renewable energy power stations are addressed. The framework and recommendations for this implementation are proposed. Data from a decentralized smart grid data system consisting of 60,000 instances and 12 attributes was used to predict the stability of the system through three different machine learning methods. The results of fitting the penalized linear regression model show the accuracy of 96% for the regression model implemented using 70% of the data as a training set. Using the random forest tree model showed as well 84% accuracy and 78% accuracy for the decision tree model and 87% for the conventional neural network.
... The new technological development has steered from adopting mechanical systems and the support to manufacturing procedures to today's highly computerized assembly lines in response to the current strong market demands and requirements. Under the Industry 4.0, the notable growth in Organization Innovation and adoption of IT has gradually influenced the consumers' perception of the product, process innovation and quality (Lee et al., 2014). ...
Article
Full-text available
The manufacturing sector of Malaysia is an essential part of its economic system. Malaysia holds a remarkable position in the global market through the exports of manufacturing products. However, this sector faces challenges such as organizations’ capability to innovate due to rapid technological and business environment changes in line with Industry 4.0. These challenges may prevent manufacturing firms from embracing Industry 4.0 and thrive in their businesses. Therefore, this research aims to identify the role of organizational practices such as Knowledge Oriented Leadership, Electronic Human Resource Management, and Decentralized Organization Structure in support of Organization Innovation to prepare manufacturing firms for Industry 4.0. A quantitative method was applied to 218 samples collected from Malaysian manufacturing firms. The results of this research are interesting for keen researchers and helpful for decision-makers and practitioners to successfully pave the way for Industry 4.0 implementation in manufacturing firms by developing innovation capabilities.
... Current competitive scenarios in industry push companies to implement modern technologies for securing their position among competitors. In this competitive scenario of the fourth industrial revolution, named Industry 4.0, new trends emerged [1], involving e.g., advanced robotics and big data analytics, which are transforming the whole manufacturing sector. The introduction of the Cyber-Physical Systems (CPS) has supported the integration of capabilities from humans and machines. ...
Article
Modular robot manipulators promise to enhance current automation practice by providing higher flexibility and quick recovery in case of failures with respect to fixed-structure robots. The configuration of the modular robot control for each new assembly, to account for the new system kinematics and dynamics, can be time consuming and requires robot modelling expertise. We propose an approach for automatically generating the Unified Robot Description Format (URDF) file of modular robot manipulators, starting from the kinematic and dynamic descriptions expressed following the URDF of the single modules they can be composed of. The approach has been implemented and numerically verified by exploiting off-the-shelf software tools from Robot Operating System (ROS) libraries.
Book
Full-text available
This book is about JIDOKA, a Japanese management technique coined by Toyota that consists of imbuing machines with human intelligence. The purpose of this compilation of research articles is to show industrial leaders innovative cases of digitization of value creation processes that have allowed them to improve their performance in a sustainable way. This book shows several applications of JIDOKA in the quest towards an integration of human and AI within Industry 4.0 Cyber Physical Manufacturing Systems. From the use of artificial intelligence to advanced mathematical models or quantum computing, all paths are valid to advance in the process of human–machine integration. The editors would like to thank the different authors for their contributions. Special thanks are extended to the Escuela Te ́cnica de Ingenieros Industriales de la Universidad Polite ́cnica de Madrid, Spain, and the Hochschule Heilbronn, Germany.
Article
Full-text available
Resumo Introdução: A Indústria 4.0 é um novo paradigma produtivo que explora a rede, a informação e a digitalização dos processos de produção, fenômeno que promete ser a quarta revolução industrial, tamanho impacto social anunciado. Objetivo: Aproximar epistemologicamente a Ciência da Informação ao fenômeno da Indústria 4.0 e verificar se as pesquisas realizadas na área são uma tendência relacionada. Metodologia: Pesquisa bibliográfica exploratória em que foram recuperados artigos científicos de bases de dados internacionais com os termos que retratam a Indústria 4.0. Resultados: Os resultados mostram que existem muitas pesquisas realizadas em Ciência da Informação e Ciências Sociais que compõem o fenômeno da Indústria 4.0 e que existe uma tendência de crescimento da produção científica a área de Ciência da Informação sobre todos os temas que compõe a Indústria 4.0. Conclusões: Conclui que existe uma tendência de pesquisa e quiçá uma mudança epistemológica em curso dentro da Ciência da Informação, possivelmente influenciada pelo novo paradigma produtivo chamado de Indústria 4.0. Abstract Introduction: The Industry 4.0 é a new productive paradigm the explore the networks, information and digitalization of productive process, phenomena that promise be the four industrial revolution, by the expected impact. Objective: epistemologically approach between Industry 4.0 and Information Science and verify if the researches realized in Information Science. Methodology: Exploratory bibliographic research in which scientific articles were retrieved from national and international databases, with the terms that report the Industry 4.0. Results: The results show that there are many researchers conducted in Information Science and Social Sciences that make up the phenomenon of Industry 4.0 and that there is a tendency of growth on all the related themes. Conclusions: It concludes that there is an epistemological change in Information Science, possibly influenced by the new productive paradigm called Industry.
Chapter
Over the past few decades, ‘digital technology’ has shaped the so-called Third Industrial Revolution—the first in the nineteenth century being characterized by steam and water, and the second at the beginning of the twentieth century being based on electricity and the emergence of mass production. In his book, The Fourth Industrial Revolution, Klaus Schwab, Founder and Executive Chairman of the World Economic Forum, suggests that it will be a further step in human production based on a complete integration between the cyber and physical dimensions. The fourth revolution has the potential to transform not only the way we produce and distribute things but also the dynamics of customer engagement, value creation, management, and regulation (Kagermann et al., 2013; Schwab, 2017a, 2017b). An historical account of the origins, history, and impact of cybernetics is beyond the scope and goals of this contribution (Ampère, 1843; Wiener, 1948a, 1948b; Simon, 1968). However, the idea of the new cyber-physical revolution or ‘Industry 4.0’ has been introduced, inspired by the transformations made in German manufacturing (Kagermann et al., 2013). Industry 4.0 has also been described as digital manufacturing, industrial Internet, smart industry, and smart manufacturing (Hermann et al., 2016; Nuccio & Guerzoni, 2019).
Chapter
This chapter explores the main markets for the production and use of robots and 3D printing technologies and presents a comparative analysis of the core robotics and 3D printing competence in the major world digital manufacturing sectors. We focus specifically on Piemonte and its efforts to develop CPS in the automotive sector, a traditional key driver of Italian industrial development. Particular attention is paid to the role of collaborative robots compared to the more traditional manufacturing robots already used heavily in automotive production.
Chapter
Full-text available
Nowadays, energy has become a key issue in all sectors of industry. Analytical tools in the prognostics and health management (PHM) area are needed to transform energy and related data into actionable information. The IMS Center has developed corresponding solutions for energy generation, storage, and usage applications. The Watchdog Agent® toolbox techniques are applied in a systematic way in each application to address the development of advanced predictive tools for near-continuous uptime of energy generating assets; mobility readiness and safety for next-generation electric vehicles via the Smart Battery Agent; and the application of low-cost, nonintrusive predictive solutions using equipment energy consumption. In this paper, methodologies and case studies in the three categories are presented.
Conference Paper
Full-text available
In present research, acoustic signals of MF285 gearbox are used for fault diagnosis of gears. Two common gears faults are used: Worn tooth face and broken tooth. Signal processing is done by discrete wavelet transform using Db4 mother wavelet. Decomposition is done in fourth level and then data mining techniques are implemented to extract several features from wavelet outputs. These features are used as inputs for artificial neural network. Feed forward back-propagation networks are used for fault classification. Two layer networks with variable neurons count in hidden layer are trained and maximum 100% performance is gained by a network with two neurons in hidden layer.
Article
Full-text available
Reliability has always been an important aspect in the assessment of industrial products. By development of technology, cost of time-based preventive maintenance increased thus, new approaches in maintenance such as condition-based maintenance (CBM) developed 1 . Machine condition monitoring has long been accepted as one of the most effective and cost-efficient approaches to avoid catastrophic failures of machines 2 . Precise and high accuracy assessment of machinery condition results in fewer stoppages and better product quality and reduce maintenance costs for plants. Thus, they can optimize workforce and implement more efficient operations 3 . Rotating machineries are used considerably utilized in the manufacturing of industrial products. Gearboxes as a key rotating motion transmission component, plays a critical role in industrial applications. Therefore, attracts research interests in condition monitoring and fault diagnosis of this equipment. Importance of gears and bearings in condition monitoring of the machine is undeniable, thus, processing and analysis of acoustic and vibration signals of the gearbox gears and bearings is the common way of extracting reliable representative of the gearbox condition 4,5,6 . Many signal processing techniques have been used for fault diagnosis, so far. Among this time, FFT is widely used in fault detection purpose. Unfortunately, FFT based techniques can not result in proper results for non-stationary signals, and by spotting that these non-stationary behaviors which could be created by environmental causes, contain reach information about machine faults. Therefor, it is important to analyze the non-stationary signals. Due to disadvantages of FFT analysis, Time-Frequency analysis methods are the most popular techniques for processing non-stationary signals. Some of common time-frequency methods are Wigner-Ville distribution (WVD), Short Fourier Transform (STFT) and Wavelet Transform (WT). Artificial intelligent systems are widely and successfully used for classification and fault diagnosis. These techniques such as Artificial Neural Networks (ANN) and Fuzzy logic are developed to mimic humans in decision-making and recognition. Due to this development, these AI methods are investigated in several researches. Kong and Chan used a combined method for fault diagnosis of triplex pump based on wavelet transform, fuzzy logic and neural networks 7 . In 2009, Chang et al, implemented probability neural networks for machinery fault classification 8 . In 2010, Jian-Da Wu and Jian-Ji Chan, identified faulted gear in a gear-set by using two types of neural networks 9 . Omid, 2010, used Fuzzy logic for a sorting and classification purpose 10 . In present study, acoustic signals of MF285 gearbox are decomposed by discrete wavelet (Db4) in fourth level. Twenty-five functions are applied into approximation and details coefficient for data mining. Processed data are fed into neural networks with different topologies where this variation are made by using several neurons count in hidden layer. Experimental Setup The experimental setup consists of a Massey Ferguson 285 gearbox mounted on a test bed and 4kW two-pole three-phase electromotor that provides drive power using belt transmission. The three-speed transmission gearbox has 1st, 2nd, 3rd and reverse crash type gears and no synchro-meshed gears included. Three states of health are studied for first gear. Following states are faults classes that are initialized on instances of gears: "Worn" tooth face and "Broken" states Broken state is prepared by removing over 50% of tooth body and Worn class by wearing tip of tooth head. Every single instance of gears is mounted inside gearbox and its acoustic emission signals acquired while gearbox operating in target gear. Data Acquisition Three different condition gears are installed inside the gearbox simultaneously and system is set to drive on 1300 RPM.
Article
Full-text available
The globalization of the world’s economies is a major challenge to local industry and it is pushing the manufacturing sector to its next transformation – predictive manufacturing. In order to become more competitive, manufacturers need to embrace emerging technologies, such as advanced analytics and cyber-physical system-based approaches, to improve their efficiency and productivity. With an aggressive push towards “Internet of Things”, data has become more accessible and ubiquitous, contributing to the big data environment. This phenomenon necessitates the right approach and tools to convert data into useful, actionable information.
Article
Full-text available
In this paper the frequency domain vibration signals of the gearbox of MF285 tractor is used for fault classification in three class: Healthy gear, Worn tooth face and broken gear. The effect of applying statistical parameters to signals on accuracy is studied. In addition, Influence of feature selection using Improved Distance Evaluation on classification performance and training speed is another target of present research. Two classification methods are used; Artificial Neural Network with variable neuron count for hidden layer in 2 layer network and k-nearest neighbor with variable k number. Using variable settings for classifier is due to make effect of statistical parameters and IDE independent from classifier settings. Results show that, accuracy improved from 86.6% to 100% by applying statistical parameters and 100% and 95.5% performance gained by applying IDE on ANN and kNN simultaneously but influence of IDE on kNN was higher than ANN.
Article
Full-text available
Cyber-Physical Systems (CPSs) are characterized by integrating computation and physical processes. The theories and applications of CPSs face the enormous challenges. The aim of this work is to provide a better understanding of this emerging multi-disciplinary methodology. First, the features of CPSs are described, and the research progresses are summarized from different perspectives such as energy control, secure control, transmission and management, control technique, system resource allocation, and model-based software design. Then three classic applications are given to show that the prospects of CPSs are engaging. Finally, the research challenges and some suggestions for future work are in brief outlined.
Conference Paper
Full-text available
This paper presents a similarity-based approach for estimating the Remaining Useful Life (RUL) in prognostics. The approach is especially suitable for situations in which abundant run-to-failure data for an engineered system are available. Data from multiple units of the same system are used to create a library of degradation patterns. When estimating the RUL of a test unit, the data from it will be matched to those patterns in the library and the actual life of those matched units will be used as the basis of estimation. This approach is used to tackle the data challenge problem defined by the 2008 PHM Data Challenge Competition, in which, run-to-failure data of an unspecified engineered system are provided and the RUL of a set of test units will be estimated. Results show that the similarity-based approach is very effective in performing RUL estimation.
Article
Battery algorithms play a vital role in hybrid electric vehicles (HEVs), plug-in hybrid electric vehicles (PHEVs), extended-range electric vehicles (EREVs), and electric vehicles (EVs). The energy management of hybrid and electric propulsion systems needs to rely on accurate information on the state of the battery in order to determine the optimal electric drive without abusing the battery.In this study, a cell-level hardware-in-the-loop (HIL) system is used to verify and develop state of charge (SOC) and power capability predictions of embedded battery algorithms for various vehicle applications. Two different batteries were selected as representative examples to illustrate the battery algorithm verification and development procedure. One is a lithium-ion battery with a conventional metal oxide cathode, which is a power battery for HEV applications. The other is a lithium-ion battery with an iron phosphate (LiFePO4) cathode, which is an energy battery for applications in PHEVs, EREVs, and EVs.The battery cell HIL testing provided valuable data and critical guidance to evaluate the accuracy of the developed battery algorithms, to accelerate battery algorithm future development and improvement, and to reduce hybrid/electric vehicle system development time and costs.
Article
For decades, researchers and practitioners have been trying to develop and deploy prognostics technologies with ad hoc and trial-and-error approaches. These efforts have resulted in limited success, due to the fact that it lacks a systematic approach and platform in deploying the right prognostics tools for the right applications. This paper introduces a methodology for designing a reconfigurable prognostics platform (RPP) which can be easily and effectively used to assess and predict the performance of machine tools. RPP can be installed on the equipment and it has the prognostic capabilities to convert the data into performance-related information. The equipment performance information can then be integrated into the enterprise asset management system for maintenance decision making through the Internet. Two industrial cases are used to validate the effectiveness of applying the RPP for different prognostic applications as well as the reconfigurable capabilities of the proposed RPP.