Applied Materials Today 10 (2018) 127–132

Contents lists available at ScienceDirect

Applied Materials Today

journal homepage: www.elsevier.com/locate/apmt

Review

Materials 4.0: Materials big data enabled materials discovery

Rajan Jose a,∗ , Seeram Ramakrishna b,c,d
a
Nanostructured Renewable Energy Materials Laboratory, Faculty of Industrial Sciences & Technology, Universiti Malaysia Pahang, 26300 Kuantan, Malaysia
b
Department of Mechanical Engineering, National University of Singapore, Singapore
c
Institution of Engineers, Singapore
d
SPRING Singapore, Singapore

a r t i c l e i n f o a b s t r a c t

Article history: Materials discovery is an incessant process and has been the landmark of human progress. This article
Received 23 September 2017 sees the evolution of materials discovery in generations, its current generation as the fourth paradigm
Received in revised form of materials research and term it as Materials 4.0, briefly describe the built-up infrastructure for the
29 December 2017
Materials 4.0, and cite few examples of materials discovery and lifecycle assessment under this protocol.
Accepted 30 December 2017
© 2017 Elsevier Ltd. All rights reserved.

Keywords:
Materials artificial intelligence
Materials informatics
Materials waste management
Materials cyberinfrastructure
Materials genomics

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
2. Evolution of materials research – an analogy with industrial revolutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
3. Data are the primary resources for Materials 4.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.1. Materials big data informatics has been set up throughout to facilitate Materials 4.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
4. Materials 4.0 on the go . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
4.1. Prediction of properties, design, and discovery of novel materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
4.2. Life cycle assessment and materials recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
5. Conclusions and outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132

1. Introduction optical susceptibility, electrical materials from insulators to super-

conductors through semiconductors, di-, piezo-, ferro-electrics and
Materials discovery lies at the heart of human progress – new so on. Significant progress has been made in engineering and
materials with high and unprecedented functions and properties manufacturing of functionally graded materials, in which tailored
as well as understanding their relationship with chemical consti- properties have been achieved through a spatial gradation in struc-
tution have been the milestones of human progress. Materials R&D ture or composition or both [1]. Our continued efforts to deploy a
have lead us not only understanding the difference in the origin minimum amount of materials for a given function lead us con-
of properties of the materials but also an exciting material-box centrating on nanostructured materials; where the current focus
of spectacular diversity as metals and alloys, high strength and is mainly on materials of flat two-dimensional network with an
conducting polymers, nonlinear optical materials of higher order atomic-layer thickness. In parallel, we have developed ability to
design materials of target functionality and simulating the pro-
cesses taking place in devices and interfaces computationally using
the principles of quantum mechanics and other materials theories.
∗ Corresponding author.
All these efforts through centuries gathered documented infor-
E-mail addresses: rjose@ump.edu.my (R. Jose), seeram@nus.edu.sg
(S. Ramakrishna). mation on materials processing, characterization, modeling, and

https://doi.org/10.1016/j.apmt.2017.12.015
2352-9407/© 2017 Elsevier Ltd. All rights reserved.
128 R. Jose, S. Ramakrishna / Applied Materials Today 10 (2018) 127–132
properties on diverse range of materials – expression of their size
requires the largest units of data quantification such as Petabytes,
Exabyte, Zettabyte, Zottabyte, or Brontobyte. Besides, our ability
to process data has increased phenomenally over the last decade
(many smartphones are more powerful than a Cray-2 supercom-
puter of 1985 that occupied a 1500 square feet room) – the Stratics
10 microprocessor family has 30 billion transistors implemented
using Intel’s 14 FinFET process and supports up to 5.5 billion logic
elements [2]; the race is now reducing the size below 10 nm thereby
significantly increasing transistor and logic gate counts, i.e., infor-
mation processing speed is increasing beyond the predictions of
Moore’s law. Now, in the context of industry 4.0, where manufac-
turing is to be undertaken employing facilities and opportunities
at the cyber physical space, the significant question to the materi-
als R&D community is “how the brontobyte data existing on the
materials processing, characterization, modeling, and properties
on diverse range of materials could be utilized in materials R&D
such that the process minimizes time span between conceptual-
ization and commercialization, reduce cost and risk of experiments
using hazardous chemicals, and preserve precious and rare materi-
als for the eventual use?” The best answer is to employ the tools of
cyber physical space and materials informatics, initiated through
the “materials genome initiative” [3] in materials R&D – a protocol
which is now getting referred as the “Fourth Paradigm of Materials
Research” or “Materials 4.0” in accordance with the latest industrial
revolution “Industry 4.0”. This protocol, with its unprecedented
computational capabilities beyond those of human’s, is expected
to solve many materials mysteries, for example, in realizing the
much sought room temperature superconductor or materials to
enable complete energy conversion and many other materials of
ideal functionalities.
2. Evolution of materials research – an analogy with
industrial revolutions
Likewise the industrial revolution be classified as first, second,
third, and fourth, materials research also could be classified sim-
ilarly. While the first industrial revolution (Industry 1.0) on the
developments of mechanical production of water and steam pow-
ered machines marked their beginning by 18th century the research
for the enabling materials, such as metals and alloys, started millen-
nia ago (in addition to the laws of thermodynamics and motion).
About 3500 BC, a superior material, called bronze, was made by
combining naturally occurring metals copper and tin, leveraging a
major technological shift now known as the Bronze Age. In analogy
with “Industry 1.0”, let us call this era of materials research to be
“Materials 1.0”. Subsequently we have seen Iron Age and Steel Age
under Materials 1.0.
In the next industrial revolution, Industry 2.0, mass produc-
tion lines employing electrical energy has been introduced by the
Fig. 1. Process flow in Materials 2.0.
beginning of the 20th century which has been enabled by under-
standing the physics and chemistry of materials as well as their
structure – property correlations, quantum theory, electrodynam-
ics, Maxwell’s equations and so on marking a clear enhancement of
the level of industrial competency over its predecessor. Industry 2.0
created enormous jobs, opportunities, and wealth. Consequently,
more systematic materials research has been enabled – which
goes through conceptualization and systematic experimentation to
prove or disprove it, prototyping in laboratory and real life environ-
ments, testing and validation of prototypes, life cycle assessment
and eventually delivering those results to enable manufacturing of
the product (Fig. 1). Let us call this phase of materials research to be
“Materials 2.0”. Over 80% of the materials research volume today
is through this protocol – this process typically take time up to 20
years or more to find a viable concept to its market entry [4]. One of
the drawbacks of the Industry 2.0 is the long supply chain involved,
i.e., all the raw materials for a given product is to be manufactured
at various other destinations and transported to the final produc-
tion line. Indeed it creates numerous jobs but the process appears
cumbersome and lesser efficient than it could be.
By 1970s, from the revolution on communication protocols,
optical communication materials, availability of numerous sen-
sors, and information technology triggered the next industrial
revolution – Industry 3.0 in which further modernization of the pro-
duction line has been enabled through electronics and IT. Perhaps,
analogously “Materials 3.0” is computational materials science,
combinatorial materials science, and similar activities where mate-
rials of a target functionality is designed computationally by
employing principles of quantum chemistry and physics. Ab ini-
tio quantum chemical methods and density functional theory has
been landmarks of materials research, where Schrodinger equa-
tions, both time dependent and independent, have been employed
with appropriate electronic and nuclear wave functions and corre-
lation functional to study the ground and excited state properties of
molecules, clusters, crystals, polymers, and glasses and their inter-
actions. However, Materials 3.0 protocol still required tools of its
predecessor for testing, validation, prototyping and life cycle anal-
ysis before enabling a concept to be delivered for manufacturing
(Fig. 2). Nevertheless, Materials 3.0 is at a markedly higher compe-
tency level than its predecessor and its practitioners, i.e., materials
scientists, engineers, and technologists were always given a “higher
grade”. One of the many citable examples are the recent effort on
the prediction of previously unknown 18-electron ternary com-
pounds using first principle thermodynamics and their accelerated
laboratory synthesis – some of them are expected to be materials
of choice for quantum electronics and renewable energy [5].
In the Industry 4.0, manufacturing is enabled through the
cyber physical system. The cyber physical system could be defined
as a system of collaborating computational elements controlling
physical entities and are physical and engineered systems whose
 R. Jose, S. Ramakrishna / Applied Materials Today 10 (2018) 127–132
Fig. 2. Process flow in Materials 3.0.
operations are monitored, coordinated, and controlled and inte-
grated by computing and communication core. One of the best
example for manufacturing under Industry 4.0 is 3D printing –
virtually any physical object, including those of organs of living
species, can be printed by atom by atom if their chemical con-
stitution could be detailed computationally. Obviously, the next
revolution in materials research, i.e., Materials 4.0 or Materials
Big Data Informatics enabled materials research [6,7], is finding
pathways of syntheses of a target material or reverse engineer-
ing a pre-designed material from the existing vast knowledge on
materials, prototyping, testing, validation and lifecycle assessment
in a virtual laboratory; thereby significantly reducing the time
from conceptualization to manufacturing (Fig. 3). Materials 4.0
is expected to deliver tools for Industry 4.0 in terms of cheaper
sensors, embedded systems, cameras and wireless communication
interfaces necessary for faster and ubiquitous data capture and
analysis to cyber physical space enabled manufacturing [8].
3. Data are the primary resources for Materials 4.0
Food, Water, and Oxygen for the Materials 4.0 is the
“big” data existing on synthesis, processing, modeling,
Fig. 3. Diagram showing the concept of ‘web based materials big data platform’ or Materials 4.0.
129
characterization, and properties of diverse range of materials
as a function of micro/nanostructure, physical conditions of
pressure and temperature, phase transitions and performance
during the fabrication of products and its life cycle, and recycling
or handling of product upon its lifetime.
3.1. Materials big data informatics has been set up throughout to
facilitate Materials 4.0
Data curation databases, synonymously termed Materials Infor-
matics or Materials Genome, are widely available since the
beginning of the current decade, i.e., since 2011 [9]. The “Mate-
rials Data Facility” [10] at the University of Chicago and Argonne
National Laboratory, The “Materials Project” at Lawrence Berkeley
National Laboratory [3,11], the “Open KIM” project on interatomic
potentials at the University of Minnesota [12,13], the “Hierarchical
Materials Design” (CHiMaD) [14] of National Institute of Stan-
dards and Technology (NIST) housed at North western University,
Argonne National Laboratory and University of Chicago, and the
“Materials Common” at the University of Michigan [15], “Predic-
tive Integrated Structural Materials Science” (PRISMS) Centre by the
Department of Energy [16], “Novel Materials Discovery” (NoMaD)
[17] of European Union are few examples. Efforts are also under-
taken to develop protocols for annotation of the materials data
and structures on the web such that the data is maintained over
time. For example, the NIST facilitated data curation by developing
a phase-based materials database to avoid conflicts in correspond-
ing materials community by adopting Unified Modeling Language
(UML) and XML schemes. The database infrastructure is based
on NoSQL as well as other standard relational technologies. Also
adopted traditional APIs, Web APIs (REST), and data exchange facil-
ities and formats (XML, JSON, BSON) to provide flexible data access
to the web-based databases, their scalability and access to new
tools of big data analytics, machine learning and artificial intelli-
gence.
In Asia, the National Institute of Material Science of Japan
(NIMS, Japan) provides one of the largest open materials databases,
MatNavi, for informatics on polymers, metals and ceramics. The
PolyInfo of MatNavi provides an online database of structure and
130 R. Jose, S. Ramakrishna / Applied Materials Today 10 (2018) 127–132
properties for pure polymers from more than 14,000 sources.
MatWeb is another searchable online material database with data
cataloged from several manufacturers and suppliers. The Materi-
als Intelligence system from Granta Design, Cambridge, UK is a
commercial platform that integrates materials data with a vari-
ety of software tools and provides Ashby charts for different
types of materials. The US based Citrine Informatics offers pri-
vate repositories for materials data spanning from development
to manufacturing. The Citrine Informatics created a web-based
database using a machine-learning-based recommendation engine
for identifying new thermoelectric materials using a large body
of experimental thermoelectric characterization data and first-
principles-derived electronic structure data as the training set. In
other words, it provides datasets collated from multiple sources
and data-driven material design tools. The NIST has developed
phase diagram based CALPHAD (CALculation of PHAse Diagrams)
for metallic alloy materials as a part of MGI [18]. It captures both
experimental and computational data related to thermodynamics,
kinetics diffusion, molar volume, elastic properties, electrical con-
ductivity, thermal conductivity and interfacial energies. In recent
years it is being developed into OpenCalphad (OC), an informal
international collaboration of scientists and researchers interested
in the development of high quality software and databases for ther-
modynamic calculations using the CALPHAD.
The “Materials Common” by PRISM at the University of Michigan
focuses on phase equilibria–microstructures–mechanical proper-
ties of metals and alloys [19], which is being developed as a platform
for accelerating the prediction of materials phenomena. The Open
Quantum Materials Database (OQMD) and the Materials Project
(MP) utilize results from density functional theory computational
methods coupled with data analytics for screening and discovery
of promising material systems. The “NanoMine” developed by the
researchers at the Northwestern University focuses on nanocom-
posites materials informatics in order to facilitate efficient material
selection and design. Diverse stakeholders of nascent materials
informatics field such as the private entities (Citrine Informat-
ics), non-profit societies (ASM International, ASME, ASTM), and
government agencies (NIST) are shaping standards for open data
frameworks, systems, and ontologies that flexibly accommodate
data for purposes of record keeping, easy retrieval via web searches,
and analysis. In order to encourage progress via open competi-
tion, NIST and Citrine Informatics hosted Materials Data Challenge
and Materials Hackathons, respectively. It is to be noted that these
materials informatics infrastructure are in early stages of develop-
ment and not yet reached the stage of easily and flexibly meet the
needs of a variety of users and products.
4. Materials 4.0 on the go
The informatics enabled materials research by developing
appropriate machine learning protocols has been already proven
in several areas of materials research. A number of reviews have
been recently published on this topic [9,20–23], which emphasize
the evolutionary algorithms to generate new and unprecedented
materials or solving the previously unsuccessful experiments. In
the following, we list few landmark experiments on the predic-
tion, design, and discovery of a range of materials systems including
metals, composites, polymers, and nanomaterials. Fig. 4 shows the
statistics of publications and citations for papers published trace-
able using the keyword “materials informatics” in Scopus. The
field is still under infancy although publications can be tracked
before this millennium; the prime factors contributing are the
time required to switch the materials research practice to a new
paradigm, expertise on big data analytics and development of
machine learning algorithms.
Fig. 4. Trend in publications and citations traceable using the keyword “materials
informatics” as on 14 Sept 2017.
4.1. Prediction of properties, design, and discovery of novel
materials
One of the unique advantages of the machine-learning model
over traditional strategies has been demonstrated by Raccuglia
et al. [24]. They used information on unsuccessful hydrother-
mal syntheses (which they call ‘dark’ reactions) collected from
their archived laboratory notebooks and added physicochemi-
cal property descriptions to the raw notebook information using
cheminformatics techniques and used the resulting data to train
a machine-learning model to predict successful reaction. When
carrying out hydrothermal synthesis experiments using previously
untested with commercially available organic building blocks, the
machine-learning model outperformed traditional strategies, and
successfully predicted conditions for new organically templated
formation of vanadium selenites with high success rate (89%).
This work is significant as it can be extended to other promis-
ing target candidates, for example, like zeolites for gas adsorption
purposes or layered cuprates for room temperature superconduc-
tivity. The potential of materials informatics and the application
of different data science techniques, including feature selection
and predictive modeling, has been employed to predicted the
fatigue properties of steels, utilizing Mat-Navi database by Agrawal
et al. [6] and presented a systematic end-to-end framework. The
approach yielded accurate results and gave confidence to build
direct models on processing and compositions. In another effort,
the limitations of the compositional and/or structural search space
for tractability of compositions have been overcome by construct-
ing a machine learning model from a database of thousands of
DFT calculations and predicted 4500 new stable materials by scan-
ning ∼1.6 million compositions without additional input and with
six orders of magnitude less computer time than DFT [25]. In
another study, materials informatics was applied to optimize the
microstructure of Fe–Ga alloy (Galfenol) to match the target elas-
tic, plastic and magnetostrictive properties [26]. Lookman et al.
[22] applied Materials 4.0 approach to discover new NiTi-based
alloys Ti50 Ni46:7 Cu0:8 Pd0:2 Fe2:3 with the smallest thermal dissipa-
tion. The used statistical inference and adaptive design for materials
discovery with steps (a) assembling a library of crystal structures
and chemistries and (b) defining the training space with a given
number of samples and features, and (c) building an inference
model using off-the shelf pattern recognition tools, such as clas-
sifiers and regressors based on linear or kernel ridge regression,
least squares regression, decision trees, Gaussian process model-
ing or support vectors. Recently, Zhang et al. [27] demonstrated
that the state-of-the-art physics-guided data mining can provide
an efficient pathway for the next generation of materials discov-
ery by classifying the alloying behavior, i.e., whether they are
 R. Jose, S. Ramakrishna / Applied Materials Today 10 (2018) 127–132
miscible or not, for 813 binary alloy systems consisting of tran-
sition and lanthanide metals. Based on an artificial neural network
algorithm and elemental similarity they have mined data from
hundreds of experimental phase diagrams, and thousands of ther-
modynamic data sets from experiments and high-throughput
first-principles (HTFP) calculations. In another landmark paper,
neural network algorithms has been applied to predict the diameter
of electrospun wires with >90% accuracy using the data gathered
from literature [28]. The electrospinning is a process that allows
fabrication of nanomaterials of diverse morphologies on a com-
mercial scale by allowing a polymeric solution to pass through an
electric field [29,30]. Inputs of their neural network model are poly-
mer concentration, applied voltage, nozzle-collector distance and
the flow rate parameters of the process; the average diameter is
the unique output of the network.
4.2. Life cycle assessment and materials recovery
In the modern times, a larger middle income population lead
to extensive use of “use and throw” products created threats
to environment in addition to the global warming; therefore
their recovery/recycle upon the product cycle is a severe envi-
ronmental problem. Many optoelectronic, energy conversion and
storage device contain rare and precious metals in small quan-
tities, recovery of these metals upon the product’s functional
lifetime is currently an active area of research [31,32]. Conse-
quently, exploring the efficiencies of computational methodologies
have become the ultimate goals. Zhang et al. [33] reported a
web-based system for reverse manufacturing and product envi-
ronmental impact assessment considering end-of-life disposition
of electronic products including desktop computers, laptop and
server. The searchable web-based system is developed with Java
Servlet and XML (eXtensible Markup Language) which is capa-
ble of seamless integration with other systems such as CAD and
PDM. Eco-Indicator 99 was used to assess the environmental per-
formance of the product, and the material information is stored
in the web-based database. They reported that the system has
been tested by a major computer manufacturer. The web-based
platform facilitates information sharing by the manufacturers,
recyclers and government agencies. It is a multi-tier decision sup-
port and evaluation system for operations in remanufacturing and
recycling include product disassembly, product recycling, material
assessment, environmental impact assessment considering End-
of-Life dispositions, product evaluation, and product and material
information management. It resulted in reducing environmental
impacts by improved materials selection and product design, and
focusing on effective recycling processes. Moreover the compa-
nies increased profits by optimizing reverse production planning.
Cerri and Terzi suggested toolsets to apply ICT technologies to
improve the lifecycle sustainability of manufacturing in a Euro-
pean context [34]. It is comprised of two tools. One is the Life Cycle
Optimization Tool that minimizes the lifecycle costs and lifecycle
environmental impacts, according to technical constraints. The tool
enables the comparison of different technological solutions and
manage the data collected from different stations/machines. Such
ICT based tools facilitate (i) real time monitoring; (ii) automatic
updating of the database thus enabling lifecycle cost and environ-
mental evaluations; (iii) enabling the comparison between real data
and theoretical estimations; and (iv) real time insights into the
operational behavior (failure rate, availability, reliability, energy
consumption, etc.) of the system. These advances enable engineers
to innovate new materials and manufacturing systems. Lu et al.
[35] described a process based “sustainable product development”
approach which incorporates lifecycle quality, lifecycle assessment,
and lifecycle cost analyses. Sheer volumes of electronic products
and their shorter life cycles placed a spot light on Eco credentials.
131
Original equipment manufacturers (OEMs) are on constant search
for environmentally benign materials and processes, and potential
for reverse manufacturing, upcycling or reusable resources recov-
ery, and zero waste. Thousands of components in each electronic
product, and often suppliers spanning the world, OEMs resort to
web-based decision support and evaluation systems.
5. Conclusions and outlook
The materials 4.0 represents a new paradigm of materials
research that overcome human limitations in coordinating and
analyzing the knowledge so far produced on materials theory, pro-
cessing, and properties by utilizing the opportunities at the cyber
physical space. Data are the primary input for this protocol; quan-
tum mechanical and other materials property simulation software,
digitized materials data, intelligent machine learning algorithms,
neural networks, and collaborative human networks of varying
capabilities are its operational tools. Materials 4.0 considerably
reduces the time span between the concept and commercialization
of products and preserve precious metals for the end use as well
as its recovery upon completion of the product lifecycle. Materials
innovations facilitated by materials selection based on the life cycle
assessment and real time materials informatics, innovative product
designs, nanotechnology, additive manufacturing, and advanced
manufacturing would enable better products with efficient use
of materials, and reduced time and cost of materials design and
deployment. Development of suitable self-healing materials would
be a step forward [36]. Reuse, recycling, remanufacturing, cascaded
use, redesign, novel waste processing, upcycling and recovery of
resources from waste are emerging mitigating options. Life cycle
engineering is a closed loop perspective aimed at zero waste of
resources, which is often called circular economy.
Several factors pose challenges to the desired resources effi-
ciency goals. For example, there are tens of thousands of materials
with diverse functional properties, real time performance, and
eco-properties, and often used in combinations, thus leading to
huge varieties of wastes. There is a need to pool, organize, ana-
lyze, interpret, and search through vast amounts and diverse sets
of real time data on materials and various stages of materials
manufacturing. For example, material selection for new type of
batteries, and piezoelectric materials for energy recovery from
waste heat and background radiation. Materials 4.0 helps to bridge
the gap between multiscale models and multiscale experiments.
Serendipitously inexpensive sensors, faster and cheaper comput-
ing power, cloud, open source and user friendly apps, big data
analytics, algorithms, machine learning and artificial intelligence
are becoming available to empower Industry 4.0; in a nutshell,
materials 4.0 is about the materials data-driven opportunities for
the future of manufacturing. These coupled with deep knowl-
edge of diverse manufacturing processes and materials will enable
resources productivity, process efficiency, environmental perfor-
mance of products and perhaps, zero waste.
The materials data are heterogeneous and complex; and there
is a challenge to aggregate data from multiple sources and diverse
datasets into simplified and searchable as well as managing and
analyzing unstructured data to enable materials 4.0, i.e., web-based
materials research platforms are to be developed by exploring
opportunities and identifying gaps via scholarly engagements such
that they are easy to be used by the globally distributed users for
accessing and uploading data as well as intuitive and easy to use.
Intuitive and visual interfaces are necessary for distributed collab-
oration. Authenticity of data, and traceability of data to the original
contributors while beefing up the cyber security and cloud stor-
age. More importantly, the web based data infrastructure should
be scalable, capable of handling multiple dynamic queries, and
132 R. Jose, S. Ramakrishna / Applied Materials Today 10 (2018) 127–132
flexible programmable access via Apps and learning tools by other
users and developers. In other words, the database technologies
(relational, NoSQL, graph, image, and time series) will need to be
further evolved. Materials 4.0 also requires development of rele-
vant international standards and codes which provide guidance in
areas such as design, use or performance of materials, products,
processes, services and systems so that they enhance the reliability
of materials data infrastructure and confidence of the users. Materi-
als 4.0 is still under infancy; strong initiatives are required to switch
the materials research practice to this new paradigm at govern-
mental, academic, and industrial levels by expanding expertise on
big data analytics and development of intelligent machine learning
algorithms.
Acknowledgements
This work is funded by the Flagship Leap 3 Research Grant of
Universiti Malaysia Pahang (http://ump.edu.my). Mr. Zaffaruddin
Bin Hamzah of Research and Innovation Department of UMP is
acknowledged for the graphics.
References
[1] M. Naebe, K. Shirvanimoghaddam, Appl. Mater. Today 5 (Suppl. C) (2016)
223–245.
[2] http://www.gazettabyte.com/home/2015/6/28/alteras-30-billion-transistor-
fpga.html.
[3] A. Jain, S.P. Ong, G. Hautier, W. Chen, W.D. Richards, S. Dacek, S. Cholia, D.
Gunter, D. Skinner, G. Ceder, K.A. Persson, APL Mater. 1 (2013) 011002.
[4] G.J. Mulholland, S.P. Paradiso, APL Mater. 4 (5) (2016).
[5] R. Gautier, X. Zhang, L. Hu, L. Yu, Y. Lin, O.L. SundeTor, D. Chon, K.R.
Poeppelmeier, A. Zunger, Nat. Chem. 7 (4) (2015) 308–316.
[6] A. Agrawal, P.D. Deshpande, A. Cecen, G.P. Basavarsu, A.N. Choudhary, S.R.
Kalidindi, Integr. Mater. Manuf. Innov. 3 (1) (2014) 8.
[7] S. Ramakrishna, T.-Y. Zhang, W.-C. Lu, Q. Qian, J.S. Choong Low, J.H. Ronald
Yune, D.Z. Loong Tan, S. Bressan, S. Sanvito, S.R. Kalidindi, J. Intell. Manuf.
(2018) (in press).
[8] S.R. Kalidindi, D.B. Brough, S. Li, A. Cecen, A.L. Blekh, F.Y.P. Congo, C. Campbell,
MRS Bull. 41 (8) (2016) 596–602.
[9] T.T. Wong, JOM 68 (8) (2016) 2029–2030.
[10] B. Blaiszik, K. Chard, J. Pruyne, R. Ananthakrishnan, S. Tuecke, I. Foster, JOM 68
(8) (2016) 2045–2052.
[11] https://www.materialsproject.org/.
[12] D.S. Karls, Transferability of Empirical Potentials and the Knowledgebase of
Interatomic Models (KIM), PhD Thesis, University of Minnesota, 2016.
[13] E.B. Tadmor, R.S. Elliott, S.R. Phillpot, S.B. Sinnott, Curr. Opin. Solid State
Mater. Sci. 17 (6) (2013) 298–304.
[14] http://chimad.northwestern.edu/.
[15] B. Puchala, G. Tarcea, E. Marquis, M. Hedstrom, H.V. Jagadish, J.E. Allison, JOM
68 (2016) 1–10.
[16] http://www.prisms-center.org/#/home.
[17] https://nomad-coe.eu/.
[18] L. Kaufman, J. Ågren, Scr. Mater. 70 (Suppl. C) (2014) 3–6.
[19] B. Puchala, G. Tarcea, E.A. Marquis, M. Hedstrom, H.V. Jagadish, J.E. Allison,
JOM 68 (8) (2016) 2035–2044.
[20] A. Agrawala, A. Choudhary, APL Mater. 4 (2016) 053208.
[21] T.C. Le, D.A. Winkler, Chem. Rev. 116 (10) (2016) 6107–6132.
[22] T. Lookman, P.V. Balachandran, D. Xue, J. Hogden, J. Theiler, Curr. Opin. Solid
State Mater. Sci. 21 (3) (2017) 121–128.
[23] L. Ward, C. Wolverton, Curr. Opin. Solid State Mater. Sci. 21 (3) (2017)
167–176.
[24] P. Raccuglia, K.C. Elbert, P.D.F. Adler, C. Falk, M.B. Wenny, A. Mollo, M. Zeller,
S.A. Friedler, J. Schrier, A.J. Norquist, Nature 533 (7601) (2016) 73–76.
[25] B. Meredig, A. Agrawal, S. Kirklin, J.E. Saal, J.W. Doak, A. Thompson, K. Zhang,
A. Choudhary, C. Wolverton, Phys. Rev. B 89 (9) (2014) 094104.
[26] R. Liu, A. Kumar, Z. Chen, A. Agrawal, V. Sundararaghavan, A. Choudhary,
Scient. Rep. 5 (2015) 11551.
[27] R.F. Zhang, X.F. Kong, H.T. Wang, S.H. Zhang, D. Legut, S.H. Sheng, S.
Srinivasan, K. Rajan, T.C. Germann, Scient. Rep. 7 (1) (2017).
[28] C. Ieracitano, F. Pantò, P. Frontera, F.C. Morabito, Commun. Comput. Inform.
Sci. 744 (2017) 27–38.
[29] S. Ramakrishna, R. Jose, P.S. Archana, A.S. Nair, R. Balamurugan, J. Venugopal,
W.E. Teo, J. Mater. Sci. 45 (23) (2010) 6283–6312.
[30] H. Esfahani, R. Jose, S. Ramakrishna, Materials 10 (11) (2017) 1238.
[31] Y.V. Nancharaiah, S. Venkata Mohan, P.N.L. Lens, Bioresour. Technol. 195
(2015) 102–114.
[32] J. Cui, L. Zhang, J. Hazard. Mater. 158 (2) (2008) 228–256.
[33] H.C. Zhang, J. Li, P. Shrivastava, A. Whitley, M.E. Merchant, CIRP Ann. - Manuf.
Technol. 53 (1) (2004) 5–8.
[34] D. Cerri, S. Terzi, Comput. Ind. 81 (Suppl. C) (2016) 47–54.
[35] B. Lu, J. Zhang, D. Xue, P. Gu, Concurr. Eng. Res. Appl. 19 (4) (2011) 307–323.
[36] K.K. Alaneme, M.O. Bodunrin, Appl. Mater. Today 6 (Suppl. C) (2017) 9–15.
Professor Rajan Jose is currently the Dean of Research
(Technology), a member of Senate, and a Senior Professor
at the Universiti Malaysia Pahang (UMP). He super-
vises the Nanostructured Renewable Energy Materials
Laboratory in the UMP. He investigated nanostructured
perovskite ceramics for microwave and superconduct-
ing electronics during doctoral research at the Council
of Scientific and Industrial Research (CSIR), Trivandrum,
India and received PhD degree in the year 2002 for his
work on superconducting and microwave electronics. He
was employed as a scientist at the Indira Gandhi Centre
for Atomic Research (India), AIST (Japan), Toyota Tech-
nological Institute (Japan), and the National University of
Singapore (Singapore) before joining UMP. He has published over 160 papers in
Web of Science indexed journals which are cited over 6300 times with an h-index
of 43. He holds over 21 patents nationally and internationally. His current research
interests include nanostructured materials and renewable energy devices. He is a
materials engineer with most of his research on the structure–property correlation
in materials for a desired device functionality.
Professor Seeram Ramakrishna, FREng, is a Highly Cited
Researcher (Clarivate Analytics, highlycited.com) in mate-
rials science. Thomson Reuters recognized him among the
World’s Most Influential Scientific Minds. He authored and
co-authored ∼1000 peer reviewed papers which attracted
∼72,000 citations with an h-index ∼127. The Webomet-
ric placed him among the only five researchers from
Singapore with h-index over 100, and among the only
1612 highly cited researchers (h-index >100) in the world.
He received PhD from the University of Cambridge, UK,
and The General Management Training from the Har-
vard University, USA. He is an elected Fellow of UK Royal
Academy of Engineering (FREng); Singapore Academy of
Engineering; Indian National Academy of Engineering; and ASEAN Academy of Engi-
neering & Technology. He is an elected Fellow of International Union of Societies of
Biomaterials Science and Engineering (FBSE); Institution of Engineers Singapore;
ISTE, India; Institution of Mechanical Engineers and Institute of Materials, Miner-
als & Mining, UK; and American Association of the Advancement of Science; ASM
International; American Society for Mechanical Engineers; American Institute for
Medical & Biological Engineering, USA. His academic leadership includes Univer-
sity Vice-President (Research Strategy); Dean of Faculty of Engineering; Director
of NUS Enterprise; Director of NUS Industry Liaison Office; Founding Director of
NUS Bioengineering; Founding Co-Director of NUS Nanoscience & Nanotechnol-
ogy Initiative, NUSNNI; and Founding Chairman of Solar Energy Research Institute
of Singapore, SERIS. He is a member of World Economic Forum’s Committee on
Sustainable Production.