Trends in Food Science & Technology 39 (2014) 47e62
Review
Intelligent food
packaging: The next
generation
Mike Vanderroosta,*,
Peter Ragaerta,b,c,
Frank Devliegherea,c and
Bruno De Meulenaerb,c
a
Laboratory of Food Microbiology and Food
Preservation, Department of Food Safety and Food
Quality, Faculty of Bioscience Engineering,
Ghent University, Coupure Links 653, 9000 Gent,
Belgium (Tel.: D32 92649905; e-mail: Mike.
Vanderroost@UGent.be)
b
Research Group Food Chemistry and Human
Nutrition, Department of Food Safety and Food
Quality, Faculty of Bioscience Engineering,
Ghent University, Coupure Links 653,
9000 Gent, Belgium
c
Pack4Food npo, Department of Food Safety and Food
Quality, Faculty of Bioscience Engineering,
Ghent University, Coupure Links 653,
9000 Gent, Belgium
Since the beginning of the current millennium, food packaging
innovation activities have gradually expanded toward the
development of intelligent packaging. This evolution reflects
the emerging need for new and efficient ways to economize
on business processes, solve safety and quality issues through
the supply chain, and reduce product losses. The general pur-
pose of this paper is to provide an overview of ongoing scien-
tific research, recent technological breakthroughs, and
emerging technologies that offer the perspective of developing
a next generation of intelligent food packaging systems to
sense, detect, or record changes in the product, the package
or its environment.
* Corresponding author.
http://dx.doi.org/10.1016/j.tifs.2014.06.009
0924-2244/ 2014 Elsevier Ltd. All rights reserved.
Introduction
Generally spoken, innovations in food packaging aim at
improving, combining, or extending the four basic func-
tions of traditional food packaging (Yam, Takhistov, &
Miltz, 2005):

Protection. Food packaging keeps food products in a
limited volume, prevents it to leak or break-up and pro-
tects it against possible contaminations and changes.

Communication. Food packaging communicates impor-
tant information about the contained food product and
its nutritional content, together with guidelines about
preparation.

Convenience. Food packaging allows for consumers to
enjoy food the way they want, at their convenience.
Food packages can be designed toward individual life-
styles through for example portability and multiple sin-
gle portions.

Containment. Containment is the most basic function of
a package and is important for easy transportation or
handling.
As society is becoming increasingly complex, users
(food producers, food processors, logistic operators, re-
tailers and consumers) continuously demand innovative
and creative food packaging to guarantee food safety, qual-
ity, and traceability. This requires appropriate technologies
that can be integrated in food packaging. For food pack-
aging innovations to be commercially viable and success-
fully adopted by the target group, they must meet the
ever increasing regulatory requirements and have a final
beneficial outcome that outweighs the possible extra ex-
penses of adding the new technology. In addition, food
packaging innovations should also aim at decreasing the
environmental pressure by taking into account a broad
range of sustainability issues (waste prevention, efficient
use of resources, process optimization, recycle, reuse .).
Food packaging innovations should therefore not only be
discussed on the basis of their scientific or technological
contributions to the four basic functions of traditional
food packaging, but also on their general contributions to-
wards a more sustainable world in which the harmful
impact of packaging waste and food loss on our environ-
ment is reduced.
Since the beginning of the current millennium, food
packaging innovation activities have gradually expanded
48 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

toward the development of intelligent packaging. Accord-

ing to the legal definition of the EU (EC, 2009), intelligent
packaging contains a component that enables the moni-
toring of the condition of packaged food or the environment
surrounding the food during transport and storage. Intelli-
gent packaging is thus a system that provides the user
with reliable and correct information on the conditions of
the food, the environment and/or the packaging integrity.
Intelligent packaging is an extension of the communication
function of traditional food packaging, and communicates
information to the consumer based on its ability to sense,
detect, or record changes in the product or its environment.
It can be derived from Fig. 1 that the number of publica-
tions on intelligent packaging has increased more rapidly
Fig. 1. The evolution (2002e2012) of the number of publications on intel-
since 2009 compared to the period before. This trend possibly ligent packaging. Source: Google Scholar: http://scholar.google.com.
reflects to a certain extent the emerging needs resulting from
the financial crisis in 2007, which has forced companies to
packaged food, or to achieve some characteristics that cannot
search for new and efficient ways to economize on business
be obtained otherwise (Miltz, Passy, & Mannheim, 1995).
processes and to reduce losses. The development of new
Intelligent packaging and active packaging are not mutually
intelligent packaging providing continuous information on
exclusive. Both packaging systems can work synergistically
the food condition or packaging integrity is not only benefi-
to realize so-called smart packaging. Smart packaging pro-
cial for the customer, but also enables the detection of calam-
vides a total packaging solution that on the one hand monitors
ities and possible abuse through the entire supply chain, from
changes in the product or the environment (intelligent) and on
farm to fork. This undoubtedly results in a safer and more
the other hand acts upon these changes (active). Although the
efficient supply chain, reducing food loss and waste and pre-
concepts of smart and intelligent packaging are often used
venting unnecessary transport and logistics from an early
interchangeably in literature, the authors of this paper would
stage. Intelligent packaging can also contribute to improving
like to emphasize that they are not the same.
Hazard Analysis and Critical Control Points (HACCP) and
To date, three major technologies exist for realizing
Quality Analysis and Critical Control Points (QACCP) sys-
intelligent packaging: sensors (and by extension nose sys-
tems1 (Heising, Dekker, Bartels, & Van Boekel, 2014) which
tems), indicators, and radio frequency identification
are developed to
(RFID) systems (Kerry, OGrady, & Hogan, 2006). These
1. timely detect unsafe foods; technologies differ from each other not only in hardware
2. identify health hazards and establish strategies and pro- (physical composition), but also in the amount and type of
cedures to prevent, reduce, or eliminate their data that can be carried and how the data are captured and
occurrence; distributed (Heising et al., 2014). In Fig. 2, it is shown that
3. identify processes that strongly affect the quality attri- the number of publications on the application of each major
butes and efficiently improve the final food quality. intelligent packaging technology has increased each year in
the period between 2002 and 2012, indicating that the
Currently, an integrated system covering the entire food
supply chain and combining material and informational
flows into one continuous food safety and quality manage-
ment process is still nonexistent.
It is important to emphasize that intelligent packaging
should not be confused with active packaging. Active pack-
aging is an extension of the protection function of traditional
food packaging and is designed such that it contains a compo-
nent that enables the release or absorption of substances into
or from the packaged food or the environment surrounding
the food (EC, 2009). Active packaging is thus a system in
which the product, the package, and the environment interact
in a positive way to extend shelf life, improve the condition of

1
A HACCP system helps food business operators look at how they Fig. 2. The evolution (2002e2012) of the number of publications on the
handle food and introduces procedures to make sure the food produced three major technologies applied in intelligent packaging: sensors, indi-
is safe to eat. cators and RFID. Source: Google Scholar: http://scholar.google.com.
 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62
domain of intelligent packaging is rapidly expanding and
maturing. This trend is also reflected in the emergence of
an increasing number of national and European research
and development projects related to intelligent packaging
(see Table 1). Furthermore, there is an enhanced focus of
national and international packaging organizations and
renowned research centers (TNO-Holst Centre, Fraunhofer
Institute, VTT Technical Research Centre of Finland,
IMEC, Technologie-Transfer-Zentrum Bremerhaven, .)
on the research and development, implementation,
commercialization and standardization of new technologies
and processes for the production of intelligent packaging
systems.
The general purpose of this paper is to provide an over-
view of ongoing scientific research, recent technological
breakthroughs, and emerging technologies that offer the
perspective of developing a next generation of intelligent
food packaging systems that can be easily integrated in
future food packaging materials in a modular way. The
scientific and technological progress will be discussed
for each major technology (sensor, nose systems, indica-
tor, and RFID) in more detail throughout this article,
with special focus on sensors as they are considered as
the most promising and game-changing technology for
future intelligent packaging systems (Bagchi, 2012;
Kuswandi, Wicaksono, Abdullah, Heng, & Ahmad,
2011). Subsequently, a brief overview will be given of
some immature research domains that could potentially
be important in the context of intelligent packaging in
the far future (more than ten years from now). To conclude
the review, a series of measures and approaches will be
provided that should be applied to design, produce and
apply future intelligent food packaging systems in a sus-
tainable way.
Sensors
Introduction
Due to intensive research efforts during the past decade,
the horizon of applying and integrating sensors in food
packaging materials has come within reach to realize intel-
ligent food packaging systems. Despite the fact that sensors
offer the prospect to provide an alternative to the time-
consuming, expensive and destructive analytical techniques
that are currently applied to monitor a packaged food prod-
uct and its environment throughout the entire supply chain,
the following obstacles still need to be overcome to
commercially apply sensors on a large scale:
- Decreasing size and reducing rigidity (improving
flexibility).
- Reducing development and production costs.
- Increasing robustness (i.e., the ability to resist environ-
mental influences such as mechanical stress, light expo-
sure, temperature variations.) and sensitivity.
- Meeting strict legislations.
- Taking into account food safety considerations.
49
Besides traditional sensors to measure temperature, hu-
midity, pH-level and light exposure, chemical sensors
have received increasing attention the last couple of years
to monitor food quality and package integrity. The sensing
part of a chemical sensor (often denoted with the term re-
ceptor) is usually a chemical-selective coating capable of
detecting the presence, activity, composition, or concentra-
tion of specific chemical analytes or gases via surface
adsorption, resulting in a change of a certain coating prop-
erty. This change is typically being observed and converted
into a proportional output signal by the actual measuring
part of the sensor, namely the transducer. If the transducer
requires external power for measurement, the transducer is
called an active transducer. If not, it is called a passive
transducer.
Small and flexible chemical sensors are particularly
interesting to develop intelligent food packaging that is
able to monitor volatile organic compounds (VOCs) and
gas molecules (H2, CO, NO2, O2, H2S, NH3, CO2, CH4,
etc) related to food spoilage and package leaking to eval-
uate the product quality and the package integrity in for
example modified atmosphere packaging2 (MAP). This
type of chemical sensors could in time offer a valuable
alternative to the currently used cumbersome analytical in-
struments, such as fixed gas chromatographyemass spec-
trometers (GCeMS) that require breakage of package
integrity, or portable gas analyzers which are not applicable
for real-time, on-line control or large scale usage.
Despite the fact that most of the existing chemical
sensor technologies are able to detect a certain quantity
of a compound with high sensitivity, they are either too po-
wer demanding (some sensors with metal-oxide semicon-
ductor transducers require operating temperatures
between 300  C and 600  C), too large, insufficiently selec-
tive (i.e., the ratio of the sensors ability to detect what is of
interest over what is not (Mottola, 2007)), too rigid, and/or
cannot be miniaturized to be mounted on or integrated in
food packaging.
However, recent technological breakthroughs and
ongoing scientific research in the domains of printed elec-
tronics, carbon nanotechnology, silicon photonics, and
biotechnology offer the prospect of developing a new gen-
eration of sensors or redesigning conventional sensors.
These advances will now be discussed in more detail,
with special focus on how they already are or soon can
be applied in the field of chemical sensors, as these are
the most complex sensors.
Printed electronics
Printed electronics is a very rapidly emerging and rela-
tively new technology, which is expected to revolutionize
2
MAP is a packaging technique in which the air surrounding a food
product is replaced by formulated gas mixtures (mostly CO2 and N2) to
extend shelf life and product quality.
50 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

Table 1. Overview of (inter)national research and development projects that were launched during the last decade within the context of intel-
ligent food packaging, in order of launch.

Project name Funding Start Duration Coordination Project description

GoodFood EU-FP6 2004 3 years Centro Nacional de Development of a new generation of
Microlelecr

onica (CSIC-CNM) e analytical methods based on micro- and
Madrid nanotechnology solutions for the safety and
quality assurance along the food chain in the
agrofood industry, including the integration of
chemical sensors in flexible RFID-tags.
SmartPack Germany-BMBF 2005 3 years Alcan Packaging Singen GmbH Integration of passive RFID tags already
during packaging material manufacturing to
deliver information on the status of the
packaged product.
Nafispack EU-FP7 2008 3 years Spanish Packaging, Transport and Packaging solution based on active
Logistics Research Centre (ITENE) components releasing natural antimicrobials
and based on intelligent functions monitoring
the quality and safety state of the product.
AIP Competence EU-Cornet 2009 2 years Industrievereinigung f
ur The active and intelligent packaging (AIP)
Platform Lebensmitteltechnologie und Competence Platform is an EU project that
Verpackung (IVLV) e Germany has been developed as an information
exchange on AIP materials, especially for
SMEs to enable them to integrate AIP
solutions into their products and adapt to
respond to consumer demand for a high
diversity of fresh and nutritious convenience
foods.
Pasteur EU-Catrene 2009 3 years NXP Semiconductors e The Development of a new multi-capability
Netherlands wireless sensor platform. The sensor
platform is based on an intelligent RFID
package in which multiple sensor
technologies are incorporated for monitoring
environmental parameters (including gases
such as oxygen and carbon dioxide for food
packaging and ethylene for fruits and
vegetables). Future applications include
monitoring boxes of perishable goods along
the supply chain.
IQ-Freshlabel EU-FP7 2010 3.5 years Technologie-Transfer-Zentrum Promoting the implementation of novel smart
Bremerhaven labels through investigation of consumer,
retailer and industry expectations.
Developing a novel smart label to monitor
temperature abuse of frozen foods, and
developing a novel smart label to monitor
oxygen content in modified atmosphere
packaged products.
FlexSmell EU-FP7-ITN 2010 3 years University Aldo Moro in Bari e Design, investigation and realization of
Department of Chemistry e Italy printed chemical sensing systems on flexible
plastic substrates for wireless compatible
applications. Applications are seen in
several fields ranging from logistic to
security. This project focused on very low-
cost and ultra low-power smart chemical
sensing tags for volatile chemicals detection
based on radio frequency identification
(RFID) for food freshness & quality
traceability control through packaging.
Lotus EU-FP7 2010 3 years Nederlandse Organisatie voor Providing the materials and the technology
Toegepast Natuurwetenschappelijk needed to produce the electric wiring in the
Onderzoek (TNO) e The flexible large area electronics by addressing
Netherlands in an integrated approach both the common
needs and the specific requirements of the
most representative applications: RFID
antennas, flexible OLED panels and flexible
photovoltaic panels.
 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62 51

Table 1 (continued )
Project name Funding Start Duration Coordination Project description
Flexibility EU-FP7 2011 4 years Technische Universitaet Dresden e Development of multifunctional, ultra-
Germany lightweight, ultra-thin and bendable Organic
and Large Area Electronics (OLAE) systems,
including disposable and rechargeable
batteries, solar cells, temperature sensors,
and RF receiver circuits. FLEXIBILITY is a
platform to gain insights regarding important
interfaces between components and
c o r r e s p o n d i n g fl e x i b l e p a c k a g i n g
technologies as enabler for complex OLAE
systems.
FoodSniffer EU-FP7 2012 3 years National Center for Scientific Development of a spectroscopic chip
Research Demokritos e Greece identifying harmful substances (mycotoxins,
allergens and pesticides) in fresh produce at
the point-of-need. FOODSNIFFER is field-
deployable and the result of the integration of
three major innovations: silicon photonics,
wafer-scale microfluidics and filtration
systems, low-power reader controlled by a
smartphone.
IsaPack EU-FP7 2012 3 years United Kingdom Materials A flexible sustainable, active and intelligent
Technology Research Institute technology platform for the packaging of fresh
food produce targeting extended shelf life and
quality, enhanced safety and reduced food
and packaging waste. Whilst suitable for a
wide range of foods, IsaPack will validate the
resulting materials and technologies for
modified atmosphere and stretch wrap
packaging of fresh meat.
PhotoSens EU-FP7 2012 3 years VTT Technical Research Centre of D e ve l o p m e n t o f a l ow - c o s t , m a s s -
Finland manufacturable, nano-structured, large-area
multi-parameter sensor array using Photonic
Crystal (PC) and enhanced Surface Enhanced
Raman Scattering (SERS) methodologies for
environmental and pharmaceutical
applications.
Smart Sensor Private 2012 NA Thin Film Electronics ASA; Bemis Flexible sensing platform for the packaging
Label Company Inc market that can collect and wirelessly
communicate sensor information monitoring
and recording key physical properties and
environmental data in packaged perishable
products.
SusFoFlex EU-FP7 2012 3 years Oulu University e Finland Novel intelligent food packaging solutions for
enhanced shelf life of food products and
protection against microorganisms. Fresh fruit
and vegetables are the initial targeted focus
with extension to meat and fish possible in
the future.
BioFos EU-FP7 2013 3 years Institute of Communication and Development of a high-added value, reusable
Computer Systems e National biosensor system for detection of food
Technical University of Athens e contamination based on optical interference
Greece and lab-on-a-chip (LoC) technology.
Promising concepts from photonics, biology
and nanochemistry will be combined, aiming
to overcome limitations related to sensitivity,
specificity, reliability, compactness and cost
issues.
CheckPack Belgium-IWT 2013 4 years Ghent University e Belgium Development of a silicon photonic based
chemical sensor for detection of food spoilage
and check package integrity of modified
atmosphere packages.
52 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62
Fig. 3. Example of a printed electronic device.
the production of electronic devices (RFID tags, displays,
sensors, batteries.) on flexible substrates (polyimide,
PEEK, PET, transparent conductive polyester, steel and
even paper) using electrically functional inks (see Fig. 3).
The market for flexible, printed electronic devices is rapidly
growing. In 2011 estimated at over 2 billion US dollar, the
global market is expected to increase to a 45 billion US dol-
lar market by 2021 (Das & Harrop, 2011). Compared to
complex traditional silicon based production processes
(with material throughputs around 0.1 m2/s), printed elec-
tronics comprises relative simple processes (roll-to-roll
manufacturing3 with throughputs up to 60 m2/s) which
are less time consuming and have less material waste.
The most commonly applied printing techniques in printed
electronics are gravure, ink-jet, and screen printing, each
having its benefits and limitations, depending on the pur-
pose. Inkjet printing is for example more applicable for
research and development activities, while gravure printing
has more potential to become a means of mass production
of printed electronics (Clark, 2010).
Today, manufacturers gradually start producing some
conventional electronic devices (e.g. amorphous silicium
photovoltaic cells, temperature sensors, .) via flexible
printing, to reduce costs. Very recently, Thin Film Elec-
tronics ASA announced that it has successfully demon-
strated a stand-alone, integrated printed electronic
temperature-tracking sensor system powered solely by bat-
teries, designed for monitoring perishable goods. The com-
pany expects that it will commercialize its new Smart
Sensor Label at the end of 2014 (ThinFilm, 2013).
Notwithstanding the fact that flexible printed electronic de-
vices have a bright future ahead, the problem of lower per-
formance is still an issue that needs to be solved. If the
performance of printed and flexible sensors can be
increased in the near future, they could in time replace or
3
In a roll-to-roll manufacturing (discrete or integrated) process, consec-
utive printing and drying processes are carried out on a moving flexible
substrate film.
be a valuable alternative to non-printed sensors due to their
unique properties:
- lightweight, bendable, rollable, portable, and potentially
foldable
- large area, thin and lower profiled
- possibility of creating sensors on a variety of substrates,
each shaped and individually tailored to operate
uniquely.
Flexible printed chemical sensors, composed of a recep-
tor printed on top of a printed transducer, undoubtedly have
the potential to revolutionize the development and produc-
tion of intelligent packaging. Although current state-of-the-
art printing techniques and manufacturing processes
already allow large-scale production of some printed elec-
tronic devices, major challenges in the development of
printed chemical sensors remain:
- The determination of optimal receptor or electronic ink
formulations that together exhibit high sensitivity and
high selectivity in the range of environmental conditions
that can be encountered during the transport and storage
of food products.
- Improving the robustness of the electronic circuits and
minimalize the power requirements of the sensor.
- Reducing intolerable variations in the production pro-
cess by identifying and improving influential process pa-
rameters (printing speed, drying, .) and material
characteristics (ink viscosity, ink deposition, smoothness
and wettability of substrate surface, .)
These challenges have resulted in a vast increase of
research efforts on this matter during the last couple of
years, reflecting the great interest in printed (chemical) sen-
sors of both researchers and the industry, in particular the
packaging and textile industry. The latter is looking for
new ways to incorporate chemical sensors in clothing to
assist workers in potentially hazardous conditions
(Monereo et al., 2011). Based on the recent evolutions, it
is therefore reasonable to expect that printed chemical sen-
sors will be ready for commercial production in the not so
far future.
Carbon nanotechnology
In the last few years there has been an exponential in-
crease in the number of published papers dealing with
nanomaterials for sensing purposes, in particular gas
sensing. This is the result of an increasing need for small,
simple, high sensitive, selective and reversible (i.e., allow-
ing multiple consecutive measurements) chemical sensors
with low limits of detection (LOD) and low operating tem-
peratures, suited for gas detection in a wide spectrum of ap-
plications (biomedical, food, warfare, .). In particular
carbon nanomaterials (CNs) such as nanoparticles (carbon
black and fullerenes), graphene, graphite (i.e., stacked
 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62
graphene) nanofibers, and nanotubes have been attracting a
great deal of research interest. Moreover, due to the ever
increasing computational power and the fact that CNs are
low-dimensional and fairly simple structured materials,
rapid numerical simulations can be carried out to predict
the physicochemical behavior of CNs for various
conditions.
CNs offer a high specific surface area and therefore
exhibit excellent detection sensitivity. Together with their
excellent electrical properties (high current density, high
electrical conductivity) and mechanical characteristics
(light weight, highly flexible, even under low temperature),
it makes them suitable to be applied in chemical sensors, as
a receptor, a transducer, or both.
Carbon nanotubes (CNTs) show concentration-
dependent changes in resistivity (Schedin, Geim,
Morozov, Hill, & Blake, 2007) due to adsorption of gas
molecules. In laboratory conditions, CNTs have shown
high sensitivity with detection limits at ppb levels. Gra-
phene exhibits similar electrical characteristics as carbon
nanotubes. Since graphene is a two dimensional material
composed of one layer of atoms, every atom of graphene
may be considered a surface atom and as a result every
atom site may be involved in the gas interactions. This
feature of graphene could in theory result in the lowest
achievable detection capability, namely a single molecule.
The relative ease of functionalization of graphene can
possibly lead to a solution to the selectivity issue of CNs-
based chemical sensors.
Fullerenes are extremely strong molecules, able to resist
great pressuresdthey will bounce back to their original
shape after being subject to over 3000 atm (Eulises,
2013). For chemical sensors, great attention is paid to the
application of fullerene and fullerene-based films as recep-
tors. A lot of ongoing research is concerned with the inves-
tigation of adsorption properties of such films towards
various organic and inorganic compounds (Grynko,
Burlachenko, Kukla, Kruglenko, & Belyaev, 2009).
Carbon nanofibers (CNFs) have received growing atten-
tion and are extensively studied. They exhibit interesting
surface properties (purity, mechanical strength, high
geometrical surface, etc.) which facilitates functionaliza-
tion and surface modification, hereby making them suitable
to be applied as a receptors.
Notwithstanding that CNs are very promising for devel-
oping a new generation of miniaturized chemical sensors
with superior performance (in the ideal case, graphene
should be capable of detecting a single molecule), their
commercial exploitation is still a way off because some ma-
jor technological barriers need to be overcome:
- Obtaining better specificity and avoiding the presence of
unwanted contaminants at the surface of the CNs. This
can be realized through surface functionalization, i.e.,
functionalizing the surface with specific chemical mole-
cules or biological components.
53
- Finding cost-effective, scalable production methods for
CN-based chemical sensors that retain the essential proper-
ties of CNs (Llobet, 2013). Some important breakthroughs
have been realized in this regard during the past two years.
In 2012, it was shown for the first time that carbon nano-
tubes could be printed (via inkjet printing) on PET and pa-
per to produce chemical sensors to detect Cl2 and NO2
vapors at sub-ppm concentration levels (Ammu et al.,
2012). Very recently, a simple and versatile rapid prototyp-
ing method was demonstrated for fabricating selective
chemical sensors from carbon nanotubes and graphite on
the surface of paper. The method is based on mechanical
abrasion of pencils containing CNs and small molecules
that interact with specific gases. These sensors are capable
of detecting and differentiating gases and vapors at a ppm
concentration level (Mirica, Azzarelli, Weis, Schnorr, &
Swager, 2013). Abdellah et al. (2013) showed the success-
ful implementation of printed CNT-based gas sensors with
exceptionally high and immediate sensor response to NH3
and CO2.
Silicon photonics
The common characteristic of most commercially avail-
able sensors and new printed and CN-based sensor concepts
is that the transducers generate an electrical output signal: the
energy associated with the detection of a certain quantity in-
duces a change of an electrical property of the receptor. This
change is observed by the transducer and converted into a
proportional electrical output signal. Due to some major
breakthroughs in silicon photonics, the research and develop-
ment of transducers generating an optical output signal has
increasingly gained interest by the scientific community
and the industry. Compared to sensors based on electrical
transducers, sensors based on optical transducers do not
need electrical power supply and can be powered and/or
read out from a distance by using UV, visible, or IR light.
Silicon based optical transducers are composed of optical
circuits which are integrated in silicon semiconductor mate-
rial. During the last decade, in particular silicon-on-insulator
(SOI) microring resonators (MR) (i.e. circular SOI optical
waveguide structures) have been increasingly recognized
as very small (size w 10 mm) and efficient optical trans-
ducers that exhibit the unique characteristic of a high refrac-
tive index sensitivity. The working principle of such devices
is based on the optical detection of small refractive index
changes: near-infrared light coming from a remote light
source propagates through the SOI MR, hereby sensing
the refractive index changes in the receptor. This finally re-
sults in a frequency shift of the light leaving the SOI.
Recently, a proof-of-concept of a chemical sensor based
on a SOI MR has been elaborated for ammonia (NH3) gas
detection. The SOI MR was coated with a nanoporous amor-
phous silicaealumina film and demonstrated sensitive
(detection limit of 5 ppm), and reversible real-time NH3
detection at room temperature (Yebo et al., 2012). Another
54 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62
silicon photonic based optical transducer that is currently
being investigated to apply in chemical sensors is a wave-
guide absorption spectrometer. Here, mid-infrared light
coming from a remote or on-chip light source propagates
through an optical waveguide structure and is absorbed by
the molecules adsorbed at the coating surface (i.e. the recep-
tor), resulting in an optical absorption spectrum. In theory,
this could lead to very selective detection of gas molecules.
However, a working proof-of-concept of a chemical sensor
based on this mechanism is currently non-existent. Main
practical challenges for implementation are noise sensitivity
(optical absorption of the coating material), low detection
sensitivity due to low gas densities and very small gas ab-
sorption cross sections (typically in the order of 1022e10
18
cm2/molecule) and the high development and opera-
tional costs of infrared lasers and detectors needed for sensor
read-out (Baets et al., 2013; Nitkowski, Chen, & Lipson,
2008; Puyol et al., 1999).
In CheckPack, a Flemish strategic basis research project
that was launched in November 2013 (CheckPack, 2013), a
very small silicon photonic based chemical sensor will be
developed to measure VOCs and CO2 concentrations in
the headspace of food packaging. During this four-year
project, it will also be explored how the previously dis-
cussed and other (new) optical transduction concepts can
be combined and integrated in the same sensor.
Besides some promising technological properties, sili-
con photonic based sensors have two important assets: 1/
low production costs and 2/the potential to produce on a
large scale. Indeed, the same infrastructure and methodolo-
gies can be applied as those applied in the production pro-
cesses of conventional silicon semiconductors for
electronic devices. Due to their significant cost, size, scale
and energy related benefits, silicon photonic based chemi-
cal sensors offer the prospect to become a commercially
viable alternative to the traditional technologies and could
in time be applied on a large scale in (bio)medical, environ-
mental, and food related applications.
Biotechnology
In living organisms, biological components like cells,
antibodies or enzymes work as natural sensing devices.
Recent breakthroughs in biotechnology have enabled the
isolation and purification of such components, hereby al-
lowing their integration in so-called biosensors. Biosensors
can be regarded as a subgroup of chemical sensors. The
main difference between biosensors and chemical sensors
is that the receptor contains biological components for
the detection of chemical analytes (Reyes De Corcuera &
Cavalieri, 2003): DNA, RNA, enzymes, phages, antibodies
or antigens, organelles, micro-organisms (bacteria, yeast,
fungi, plant and animal cells), biological tissue, or biomi-
metic components.4 Biosensors can be applied to identify
4 possible hazardous effects of the biological components
Biomimetic components are synthetic components, e.g. polymers, that
mimic (act like) their biological counterpart.
and measure allergens and analytes such as sugars, amino
acids, alcohols, lipids, nucleotides, etc.
Biosensors find application in medicine, pharma,
food and process control, environmental monitoring,
defence and security, but most of the market of over
US$13 billion is driven by medical diagnostics and, in
particular, glucose sensors for people with diabetes
(Turner, 2012). The latter were the first commercially
available biosensors (invented by Clark in 1962) and
allow diabetic patients to monitor their blood glucose at
home with a single drop of blood. In the development
of commercially viable biosensors, researchers always
face the following issues:
- Immobilization of biological components in the receptor
by means of physical adsorption, membrane entrapment,
covalent bonding, cross-linking, or microencapsulation
(Malaysia, 2013).
- Preventing the denaturation or degradation of the bio-
logical components due to environmental conditions
(pH, radiation, temperature, humidity, mechanical
stress).
- Finding fast, scalable and cost-effective production pro-
cesses of biological components. Biomimetic compo-
nents are more and more considered as promising
alternatives to overcome the latter. They are not only
cheaper to produce, but also lack instability and, in
some instances, irreproducibility (Turner, 2012). Unfor-
tunately, the affinity shown by these biomimetic compo-
nents is still several orders of magnitude below that of
the antibodies (Barcel
o, Rodriguez-Mozaz, & Lopez de
Alda, 2006).
- Ignoring the input generated by analytes other than the
analyte of interest or actively suppress all interactions
other than those with the targeted analytes. The latter
can be realized by selecting an appropriate biological
component that will interact only with the targeted ana-
lyte. However, in real-world conditions, the exact
composition of the environment is usually unknown
and can vary widely. Some of these unknown analytes
may also interact with the selected biological compo-
nents and hence induce a deviation in the measurement
(Barcel
o et al., 2006).
With regard to the integration of biosensors in food
packaging, extensive research efforts in the biomedical
domain during the last decade have further narrowed the
gaps between the theoretical concept of an integrated
biosensor, a working proof-of-concept and its practical im-
plementation. To date, most biosensors developed for food
industry applications are constrained to preliminary proof-
of-concept investigations (Pilollo, Monaci, & Visconti,
2013) and require further research to integrate them in
food packaging. Special attention needs to be paid to the
in biosensors on the contained food.
 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62
In 2011, Flex Alert, a research and development com-
pany of Scheelite Technologies LLC, presented a commer-
cially available flexible biosensors for detection of
pathogens in food grains, perishable fruits, and wine pro-
duction (Scheelite Technologies, 2011).
Nose systems
Food production processes (ripening, fermentation,
cooking .) or food spoilage are mostly associated with
the presence of certain flavours/odours/savours, i.e., typical
combinations of released compounds. Since individual
chemical sensors or biosensors are mostly designed to be
highly selective and/or sensitive for a limited selection of
specific compounds, a system is required that is able to
detect every compound present in the odor. Such a system
is called a nose system. A nose system mimics or exceeds
the human sense of smell or taste by generating a unique
response to each flavour/odour/savour. A nose system com-
prises a one- or two-dimensional array of chemical sensors
or biosensors with partial specificity and statistical methods
(e.g., machine learning) enabling the recognition of simple
or complex flavours/odours/savour (Gardner & Bartlett,
1993).
The ideal chemical sensors or biosensors to be integrated
in a nose system should fulfill the following criteria:
- Sensitivity similar to or better than the human nose.
- Selectivity to different compounds; compounds should
be detected by at least one sensor.
- Small dimensions, high stability, high reproducibility
and reliability.
- Short response and recovery time.
The previously described evolutions and scientific
breakthroughs in the domains of printed electronics, carbon
nanotechnology, silicon photonics and biotechnology could
eventually also contribute to the development of a new gen-
eration of cheaper and smaller nose systems. These could
then be integrated in food packaging, as a replacement of
or alternative to the current large, expensive and rather rigid
nose systems with traditional chemical sensors or biosen-
sors that are currently being used for research purposes or
in the food industry for production process control. Elec-
tronic nose systems for example cost about US $40,000
to $50,000/piece at the end of the previous century
(Ouellette, 1999).
Indicators
Introduction
In contrast with sensors, indicators cannot provide quan-
titative information about a quantity (e.g., concentrations,
temperature .) and cannot store the data of measurement
and time. They provide immediate visual, qualitative (or
semi-quantitative) information about the packaged food
by means of a color change, an increase in color intensity
or diffusion of a dye along a straight path (Kerry et al.,
55
2006). In most of the cases, the basic requirement of an in-
dicator is that the color or intensity changes or diffusions
are irreversible. If not, this may cause possible false infor-
mation (Pavelkov
a, 2012). The exhaustive research efforts
during the last decade have resulted in a vast amount of sci-
entific publications on indicators (or labels) and a growing
range of commercially available indicators for food pack-
aging (see Table 2).
Of all technologies discussed in this paper, indicators,
together with flexible RFID tags which will be discussed
in the next section, can be regarded today as the most
matured and most commercially viable technologies to be
integrated in intelligent packaging. In the following subsec-
tions, the different categories of commercially available in-
dicators will be discussed briefly.
Gas indicators
Package integrity is an essential requirement for the
maintenance of quality and safety of food products in
MAP. Gas indicators offer an alternative, non-invasive
approach to traditional destructive techniques to determine
the package integrity (e.g., leaking seals) of MAP. They
usually provide qualitative or semi-quantitative information
about altered gas concentrations (CO2, O2, water vapor,
ethanol, .) through visual colorimetric changes (Vu &
Won, 2013).
Freshness indicators
Freshness indicators provide immediate product quality
information resulting from microbial growth or chemical
changes within a food product. Microbiological quality
may be determined visually through reactions between mi-
crobial growth metabolites and integrated indicators within
the package. Freshness indicators can also be used to pro-
vide an estimate about the remaining shelf life of perishable
products (Kuswandi, Maryska, Jayus, Abdullah, & Heng,
2013).
Timeetemperature indicators
Temperature is usually the most important environ-
mental factor influencing food deterioration. Timeetemper-
ature indicators (TTIs) provide visual information of
temperature history during distribution and storage, which
is particularly useful for warning of temperature abuse for
chilled or frozen food products (Pavelkov
a, 2013).
Thermochromic ink
Thermochromic ink is a specialized dynamic ink that
changes color with exposure to different temperatures. The
color change of thermochromic inks can be either irreversible
or reversible. Irreversible thermochromic inks are invisible
until exposed to a certain temperature at which an intense co-
lor develops. Once this color progresses it will remain con-
stant or it will change colors leaving a permanent
indication of a temperature change. Reversible thermochro-
mic inks change color when heated and return to original
56 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62

Table 2. Non-exhaustive list of commercially available indicators for food packaging.

Indicator category Company or product name Information

Gas Novas Insignia Technologies A specialized pigment for use in plastic packaging which shows a
clear color change when packaging has been damaged for
products packed in a modified atmosphere. This allows
manufacturers and retailers to remove this product from the
supply chain before it even reaches the supermarket shelf.
O2Sense An eye readable indicator that gives a clear, visual indication by
means of a color change as to whether the amount of oxygen that is
present in a sealed food package is within the specified limits. This
indicator is intended to alert producers, retailers and consumers of
possible breaches of integrity in the package that could lead to an
unsafe product.
Freshness FreshTag Color indicators sensing the production of volatile amines, which
produce the familiar fishy odor that is common to all seafood.
The odor-causing chemicals react with the patented, non-toxic
food dye based indicator and gradually produces a color reaction,
indicating that the seafood is past the point of useable freshness.
RipeSense Label that changes color to indicate the ripeness of fruit. The
ripeSense label works by reacting to the aromas released by the
fruit as it ripens. The label is initially red and graduates to orange
and finally yellow.
TimeeTemperature 3M MonitorMark Blue compound run-out to estimate how long a product was above
a certain threshold temperature (ranging between 15  C
and 31  C). There are two versions, one intended for
monitoring distribution and other intended for consumer
information.
Timestrip Complete Actively monitors high and low threshold temperature breaches
outside the range 2e8  C.
Fresh-Check A self-adhesive indicator composed of an active zone that darkens
irreversibly e faster at higher temperatures and slower at lower
temperatures. It enables consumers to see when to use or not use
the food product within the product date codes.
CheckPoint Indicator adapted to toxin formation of Clostridium botulinum in
the temperature range between 1  C and 25  C conforming to the
FDA requirements of packed seafood products imported to the
USA.
CoolVu Food Over a period of time (determined by the manufacturer), an active
zone fades from silver to white. The higher the storage temperature,
the faster the fading.
Innolabel Timestrip Time and temperature sensitive label designed to provide visual
information and greater control for consumers, e.g. providing an
accurate shelf life guide for fresh fruit and vegetables rather than
a stagnant date.
Thermochromic inks LCR Hallcrest; Chromatic Technologies, Reversible color changes depending on temperature.
Inc.; Matsui International Company, Inc.

color when temperature decreases (Sarley, 2011). The activa-
tion temperatures of thermochromic inks range from low
refrigeration temperatures through human body tempera-
tures to high temperatures that exceed the pain threshold:
- Cold Activated Thermochromic Ink is used on labels and
packaging to create a color change when cooled.
- Touch Activated Thermochromic Ink will become trans-
parent when rubbed or touched to reveal an image or
another color printed beneath.
- Touch Activated Liquid Crystal Ink will change color
within the visible spectrum when rubbed or touched.
- High Temperature Thermochromic Ink is designed to
change color just below the pain threshold alerting con-
sumers and users to a safety hazard.
Thermochromic ink can be applied to realize intelligent
packaging by for example ensuring consumers that a
beverage in a drink container is perfectly chilled or to
alert consumers that a package in the microwave has
reached the desired temperature or is too hot.
Radio frequency identification (RFID)
Introduction
RFID technology does not fall into either the sensor or
indicator classification but rather represents a separate tech-
nology. RFID is grouped under the term Automatic Identi-
fication (Auto ID), together with barcodes, QR-codes,
magnetic inks, voice recognition, biometrics etc. Auto ID
technologies are a relative new way of providing
 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62 57

coupling. Both approaches can transfer enough power

Fig. 4. RFID tag.
information and/or controlling material flow, particularly
suitable for large production networks such as food supply
chains. Compared to sensors and indicators, Auto ID tech-
nologies do not provide qualitative or quantitative informa-
tion about for example the product quality status. They are
typically applied for purposes such as identification, autom-
atization, antitheft prevention or counterfeit protection
(McFarlane & Sheffi, 2003).
In a typical RFID system, a reader (i.e., a read/write de-
vice composed of a transmitter and/or a receiver) uses elec-
tromagnetic (EM) waves to communicate with an RFID tag
through antennas. The RFID tag, sometimes also denoted
as label, is a data carrying device that is composed of a
microchip attached to an antenna (see Fig. 4).
RFID tags may be classified into three types on basis of
power supply (Ilie-Zudor, Kem
eny, Egri, & Monostori,
2006):
- Passive RFID tags have no battery and are powered by
the EM waves emitted by the reader. Two fundamentally
different RFID mechanisms exist for transferring power
and data between the reader and the tag (Kaur, Sandhu,
Mohan, & Sandhu, 2011): induction and backscatter
to a remote tag to sustain its operation, typically between
10 mW and 1 mW, depending on the tag design.
Compared to other tag types, passive RFID tags are
the smallest and lightest and can be read at a range of
a few meters.
- Semi-passive RFID tags use a battery to maintain mem-
ory in the tag or power the electronics that enable the tag
to modulate the EM waves emitted by the reader an-
tenna. The battery also helps to extend the transmission
range of the data going back to the reader.
- Active RFID tags are powered by an internal battery,
used to run the microchips circuitry and to broadcast
a signal to the reader. Active tags generally ensure a
longer read range than passive tags, but are more expen-
sive than passive tags. An active tags lifetime is limited
by the stored energy, balanced against the number of
read operations the device must undergo. An active
RFID tag can typically be read at a range of 100 m.
RFID tags fall into three categories as regards to the fre-
quency of the EM waves that are used for communication.
The frequency determines the reading range and the data
transmission rate (see Table 3). Although RFID technology
has been available for many years for tracking and identifi-
cation purposes in various domains (e.g. human identifica-
tion, ticketing, sports events timing, shipping management),
the maximum exploitation of its potential for application in
intelligent food packaging systems still requires some tech-
nical, process and security issues to be solved ahead of
time. Within this context, todays research is mainly
focussed on sensor-enabled RFID tags and their beneficial
application in food supply chains. This will now be dis-
cussed in more detail.
Sensor-enabled RFID tags
Originally, RFID tags were only used to track food prod-
ucts in the food supply chain during distribution and stor-
age. RFID technology served as a replacement for
barcode scanners for this particular application. In June
2003, US largest retailers Wal-Mart and Best Buy made
the announcement to require their top 100 suppliers to pro-
vide RFID tags on pallets and cases. This decision brought
RFID into the spotlight and generated considerable

Table 3. Frequency categories and ranges that are applied for RFID communication.

Frequency category Range Information

Low (LF) 30e500 kHz Cheaper than any of the higher frequencies. Fast enough for most applications, however for large
amounts of data the required communication time will increase considerably. The disadvantage of
these frequencies is the short reading range.
High (HF) 10e15 MHz High frequencies enable higher data transmission rates and larger reading ranges but are more
expensive.
Ultra High (UHF) 850e950 MHz Largest reading range of all frequencies (3e6 m for passive tags and 30 m for active tags). Very
2.4e2.5 GHz high data transmission rates, allowing short communication times. This feature is important where
5.8 GHz tagged entities are moving with a high speed and remain only for a short time in a readers range.
58 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62
attention on using RFID in the supply chain to track and
control the movement and handling of raw materials and
products with great precision throughout several processes
along the supply chain (SCDigest Editorial Staff, 2009).
Due to recent developments and intensive research efforts,
the horizon of applying RFID technology to also monitor
storage conditions (temperature, humidity, light, .) and
food quality through the entire supply chain has come
within reach. These additional features will undoubtedly
further contribute to the successful application and adop-
tion of RFID-based devices and will continuously improve
the overall performance of supply chains (McCartney,
2006). They will become faster, more efficient and produc-
tive, more secure and better tuned to consumer preferences.
Which all translates to lower costs, better customer service,
enhanced control of stock and inventory size, better man-
agement of the cold chain for perishables, reduction of
food loss, an auditable electronic trail of events from land
to fork, increased profits and brand loyalty (Ilie-Zudor
et al., 2006).
To make RFID systems more intelligent, RFID tags
should thus also be able to provide (historical) information
about the integrity of the package, the quality status of the
food and the environmental conditions during transport or
storage. This requires the measurement of one or more of
the following properties: temperature, relative humidity,
pH, pressure, light exposure, volatile compounds and gas
molecules concentrations. This can be realized by connect-
ing one or more sensors to a RFID tag to ensure energy sup-
ply of the sensors and storage of the data measured by the
sensors in the memory of the tag (Abad et al., 2007;
Sample, Yeager, Powledge, Mamishev, & Smith, 2008).
This remote and non-destructive way of monitoring food
products is especially relevant for the transport and storage
of perishable goods, such as fruit, vegetables, meats and
fish. These products need to be kept under very strict con-
ditions to ensure freshness and quality. Sensor-enabled
RFID tags could help food distributors to optimize their
supply chain by minimizing the amount of food that is
spoiled before it reaches the retailer.
The main issues to be solved in current and future R&D
projects in the domain of sensor-enabled RFID tags are the
integration of one or more sensors in the design of RFID-
tags and the integration of (flexible) sensor-enabled RFID
tags in packaging materials. Already in 2007, the European
GoodFood project (see Table 1) demonstrated a first proto-
type of a semi-passive flexible RFID tag with integrated,
low power sensors to monitor relative humidity, tempera-
ture and light intensity.
Recently, Holst Centre in the Netherlands developed a
prototype of a flexible RFID tag with integrated sensors
able to monitor temperate and humidity and a resistive
sensor capable of detecting the presence of amines (Smits
et al., 2012). The tag is able to detect traces of trimethyl-
amine (TMA), which is a marker of interest to evaluate
the freshness of fish. The clear response to even as low
concentrations as 1 ppm suggests that the tag may be
well suited for detecting the presence of TMA in the head-
space of food packaging.
In 2013, the pan-European Pasteur project (see Table 1)
closed by announcing a technology demonstrator that inte-
grates an RFID chip, microcontroller and sensor platform,
incorporating temperature, relative humidity and light sen-
sors, into a flexible tag. In addition, two chemical sensors
have been realized as standalone demonstrators. The result-
ing oxygen sensor delivers state-of-the-art sensitivity but,
unusually, operates at room temperature, hereby reducing
power requirements. For the carbon dioxide sensor, the
research team achieved unprecedented sensitivity in the
300e5000 ppm concentration range used in food packaging
applications. These could eventually also be incorporated in
the flexible tag to monitor the controlled atmosphere in
which many foods are packaged.
It is expected that the European FlexSmell research proj-
ect (see Table 1) will soon release a prototype of a new flex-
ible sensor-enabled RFID tag with VOC and gas sensing
capabilities.
Based on the description of ongoing research projects
such as SusFoFlex and IsaPack, new sensor-enabled
RFID tag concepts are currently being investigated to
- Improve the sensitivity and selectivity of the chemical
sensors (gas and/or VOC).
- Realize the integration of sensors in RFID tags via state-
of-the-art printing techniques.
- Develop new methods to integrate sensor-enabled RFID
tags in packaging materials.
Although the current market for sensor-enabled RFID
tags is still immature and expanding, each year more and
more manufacturers of RFID tags expand their business to-
ward sensor-enabled RFID tags. Today, mountable (i.e., tag
is not an integral part of the packaging material), non-
integrated (i.e., sensors not integrated in the circuit design
of the RFID tags) and non-flexible sensor-enabled RFID
tags with sensors able to monitor the temperature, relative
humidity, light exposure, pressure and/or pH of products
are already widely commercially available (see Table 4
for a concise list). These RFID tags can be mounted on
food packaging or put in boxes or containers in which
packaged food products are transported, for example to
detect possible interruptions of the cold chain or other in-
fringements which are harmful to the food quality or
safety.
Passive RFID sensors
In the previous section, it was discussed how RFID tags
are combined with one or more (integrated) sensors which
only measure a specific aspect of the environment (temper-
ature, oxygen concentration, relative humidity, .).
Recently, a new paradigm has been proposed to apply a
simple, passive RFID tag as a sensor by using the generated
 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62
Table 4. Non-exhaustive list of several commercially available non-integrated and non-flexible sensor-enabled RFID tags.
Product name Monitoring function
Intelleflex TMT-8500 Temperature
CAEN RFID easy2log Temperature
RT0005ET
SecureRF Lime Tag Temperature. Expandable to
2.0 Sensor shock sensing, relative
humidity, and pH-level.
AMS SL13A Temperature. Expandable with
one external sensor.
EM field in the tags antenna to sense the environment,
without making use of specific designed sensors or batte-
ries. The properties of the EM field are affected by the pres-
ence of (specific components in) food products, which on
its turn induces a change in the electrical properties of
the tags antenna. A recent study demonstrated the applica-
bility of passive RFID tags as sensors to monitor different
aspects of food quality, including milk freshness, bacterial
growth, and fish spoilage (Potyrailo et al., 2012). The use
of passive RFID tags as sensors could be particularly inter-
esting when battery-free operation is critical.
Intelligent packaging technology: beyond the horizon
In this section, an overview is given of immature tech-
nologies that could potentially be applied in the context
of intelligent packaging in the far future.
Organic photonics and electronics
In organic photonics and electronics, it is currently
investigated how optical and electrical circuits can be inte-
grated in organic materials (e.g. polymers) instead of sili-
con, without compromising on the size of the circuits
currently realized with silicon (w10e100 nm) and aiming
for similar (or better) optical or electrical characteristics
and superior mechanical properties. This research of course
also encompasses the design and synthesis of new organic
materials. Because research activities in this domain are
still in an early, fundamental stage, commercial devices
are not to be expected in the near future.
Carbon photonics
At the beginning of this millennium it was found that
carbon nanomaterials (CNs) not only exhibit superior elec-
trical and mechanical properties, but that they also demon-
strate unique optical properties that could be applied in the
development of future optical sensors, as an alternative or
complement to silicon photonics. This has lead to the emer-
gence of the first research papers in the area of optical bio-
sensors based on CNs at the beginning of this decade
(Kruss et al., 2013). In particular so-called carbon dots
(CDs) have been identified as class of strongly fluorescent
and emission-color-tuning CNs with great analytical and
59
Frequency Power supply Range
902e928 MHz (North America) Semi-passive 100 m
865.6e867.6 MHz (Europe/India)
860e928 MHz Semi-passive 8m
860e960 MHz Semi-passive 20 m
13.56 MHz Passive Near field
Semi-passive communication (NFC)
bioanalytical potential (Esteves da Silva & Goncalves,
2011).
Internet of everything
The internet of everything (IoE) is a relative new concept
that aims at a world-wide network of interconnected objects
provided with sensors and RFID tags. The US National In-
telligence Council foresees that by 2025 not only mobile
phones, tablets, laptops and personal computers will be
part of the IoE, but also less obvious things like appliances,
food packages, furniture, cars, bikes, etc (Atzori, Iera, &
Morabito, 2010). This is possible through the integration
of several state-of-the-art technologies and communications
solutions, such as RFID tags, GPS, wired and wireless sen-
sors, enhanced communication protocols, etc. It is expected
that the IoE will soon be able to monitor, manage and con-
trol (real-time and/or remote) many aspects of life, such as
traffic, supply chains, logistics, personalized advertising,
environment, patient health, tracking and tracing of objects,
etc. The IoE will result in the generation of enormous
amounts of different data which have to be stored, pro-
cessed and presented in a concise and interpretable form.
New machine learning methods, data mining algorithms
and decision support systems will have to be developed
to handle such so-called big data.
With regard to intelligent packaging, the IoE comprises
more than just putting RFID tags and/or sensors on food
packaging to for example remotely monitor the quality of
the food and the integrity of the package. It is expected
that for intelligent packaging, the IoE will eventually result
in an advanced food safety management, HACCP and
QACCP systems that will be able to correctly (Takhistov,
2009):
- monitor food loss and food waste on an international
scale;
- identify potential hazards and conduct biohazard
analysis;
- recommend controls, critical limits, and appropriate
corrective actions when a deviation occurs.
Such systems will enable the monitoring and control of
flows of both materials and information in the food supply
chain cycle (Yam, 2012). The authors of this paper believe
60 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62
that the development of intelligent packaging systems re-
quires mathematical models at three different levels in the
supply chain. At the level of an individual package, math-
ematical models are needed to process raw, multidimen-
sional signals from the intelligent devices, relate them to
real world quantities (temperature, concentrations, volume,
position, .), and classify the resulting data into different
categories (e.g., food freshness level, package integrity sta-
tus). At an intermediate level (e.g., a certain batch of pack-
ages or a certain supply chain stage), mathematical models
are needed to find patterns in data (e.g., finding patterns
indicating emerging food safety threats and/or hazards).
At the level of the entire supply chain, mathematical
models are needed to make predictions and estimations
based on historical and present data (e.g., shelf life predic-
tions, food loss estimations) or to support producers, dis-
tributors and decision makers in taking appropriate and
effective actions and understanding their implications
(Yam, 2012).
Smart packaging
A major challenge for the coming decade(s) is to
develop so-called smart packaging, i.e., combining and
integrating the active and intelligent packaging concepts
in one packaging material. This kind of closed-loop pack-
aging systems offers the perspective to monitor changes
in the product, packaging, and/or the environment and to
respond appropriately on these changes via a feedback
mechanism. This will require the development of new tech-
nologies, e.g. printed electronic systems with integrated
sensors which are connected to other, yet to be developed,
devices enabling for example the release or absorption of
substances or activating certain non-thermal preservation
processes through different kinds of techniques, including
UV irradiation, gamma irradiation, ultrasound treatment,
and application of high-voltage pulsed electric fields
(PEFs) (Ortega-Rivas, 2012).
Sustainable intelligent food packaging: cradle-to-
cradle design approach
Earlier in this review, it was discussed how intelligent
food packaging can improve the overall performance of
supply chains, hereby contributing to a global reduction
of food loss and food waste. Intelligent food packaging sys-
tems should however not only be sustainable in terms of
their application, but also in terms of their design and
production.
Since the Industrial Revolution (18the19th century),
most manufacturing systems are based on a one-way flow
of materials, from cradle-to-grave (C2G). This design
approach is inherently associated with the fact that every
designed product or package eventually ends up as un-
wanted waste that must be dealt with at some cost. More-
over, since it is a one-way flow out of the factory, the
manufacturer loses the value of reusing the material. The
C2G design approach is based on taking, making and
wasting and is therefore a major burden on our environ-
ment and our health. Only 30e40 years ago governments
and industries became aware and gradually accepted that
the C2G design approach is depleting our resources (mate-
rials and fossil fuels) at a very high rate, hereby compro-
mising the future of humanity. This resulted in a change
in attitudes and the rise of so-called eco-efficiency strate-
gies to promote consumption reduction, lifespan extension
of products, prevention of waste and emissions, and reduc-
tion of the effects, without suggesting a real alternative to
the C2G material flows (Stouthuysen & le Roy, 2010).
Eco-efficiency strategies thus only focus on the reduction
of environmental impacts made by human activity
(Verfaille & Bidwell, 2000) and stand in the way of a
fundamental redesign of material flows. Typically, most
of recyclable materials are reused in lower applications
(downcycling, i.e., a downgrade in material quality) and
eventually end up as being pure waste.
In their attempt to tackle the shortcomings associated
with the C2G design approach and the eco-efficient way
of thinking, McDonough and Braungart proposed in 2003
a new design paradigm, the so-called cradle-to-cradle
(C2C) design approach supported by eco-effectiveness stra-
tegies (McDonough, Braungart, Anastas, & Zimmerman,
2003). In contrast to eco-efficiency, eco-effectiveness
directly (i.e., already during the conceptualization phase
and design process of a new product) deals with the issue
of maintaining (or upgrading) resource quality and produc-
tivity through many cycles of use, rather than seeking to
eliminate waste afterward (Braungart, McDonough, &
Bollinger, 2006). This of course requires a collaborative,
multidisciplinary and cross-sectoral approach. The
following three principles are essential in the C2C
approach:
- Waste equals food, meaning that everything is a nutrient
for something else.
- Use of energy sources that are renewable in the time-
frame they are used.
- Promotion and combination of biological, cultural, and
conceptual diversity.
Food packaging designed according to the above C2C
principles is either a nutrient that can be reused, or recycled
in a closed-loop process with zero loss in material. What
currently stands in the way of closed-loop recycling of
food packaging is that materials (foils, additives, coatings,
inks, .) and intelligent devices (sensors, RFID tags, indi-
cators, .) are not designed or selected with closed-loop re-
cyclability in mind. Today, packaging materials must be as
cheap as possible, limiting certain design aspects and often
leading to multilayer composites or laminates that are diffi-
cult or impossible to reuse or recycle. The lack of foresight
in the design of packaging and the extensive use of many
different materials inevitably results in complex and expen-
sive recycle processes, reduced performance and
 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62
attractiveness of recycled packaging, and downcycling of
packaging materials (Braungart & McDonough, 2002).
Furthermore, intelligent devices are often conceived as
relatively cheap, disposable single-use (irreversible) de-
vices because their production cost can be kept low, hereby
ignoring concepts such as reusability and/or reversibility.
To gradually introduce the C2C design approach in the
packaging industry, the Sustainable Packaging Coalition
has been established, an international consortium of indus-
try members that envisions a closed loop system for all
packaging.
Conclusion
Various emerging technologies are currently being
examined and all offer the perspective of being integrated
in new intelligent food packaging systems that meet the
needs and requirements of different actors throughout the
food supply chain. Notwithstanding that the research on
these emerging technologies is immature and a lot of hur-
dles still need to be overcome, it is expected that the next
generation of intelligent packaging will allow a better
monitoring of the flow, safety and quality of food products.
This will undoubtedly result in a better control of food sup-
ply chains (whether or not supported by smart packaging
systems and/or decision support systems), and a further
improvement of their overall performance and security.
Acknowledgments
This review paper was obtained in the framework of a
project of Pack4Food npo, supported by the Institute for
the Promotion of Innovation by Science and Technology
in Flanders, Belgium (IWT) and by a diverse group of in-
dustrial stakeholders within the packaging industry.
References
ThinFilm. (2013, October 16) Retrieved November 14, 2013, from:
http://www.thinfilm.no/news/stand-alone-system/.
Abad, E., Zampolli, S., Marco, S., Scorzoni, A., Mazzolai, B.,
Juarros, A., et al. (2007). Flexible tag microlab development: Gas
sensors integration in RFID flexible tags for food logistic. Sensors
and Actuators B: Chemical, 127, 2e7.
Abdellah, A., Abdelhalim, A., Loghin, F., K ohler, P., Ahmad, Z.,
Scarpa, G., et al. (2013). Flexible carbon nanotube based Gas
sensors. IEEE Sensors Journal, 13e10, 4014e4021.
Ammu, S., Dua, V., Agnihotra, S. R., Surwade, S. F., Phulgirkar, A.,
Patel, S., et al. (2012). Flexible, all-organic chemiresistor for
detecting chemically aggressive vapors. Journal of the American
Chemical Society, 134, 4553e4556.
Atzori, L., Iera, A., & Morabito, G. (2010). The internet of things: a
survey. Computer Networks, 54, 2787e2805.
Baets, R., Subramanian, A. Z., Dhakal, A., Selvaraja, S. K.,
Komorowska, K., Peyskens, F., et al. (2013). Spectroscopy-on-chip
applications of silicon photonics. In J. E. Broquin (Ed.), Integrated
optics: Devices, materials, and technologies XVII.
Bagchi, A. (2012). Intelligent sensing and packaging of foods for
enhancement of shelf life: concepts and applications. International
Journal of Scientific & Engineering Research, 3(10).
61
Barcel
o, D., Rodriguez-Mozaz, S., & Lopez de Alda, M. J. (2006).
Biosensors as useful tools for environmental analysis and
monitoring. Analytical Bioanalytical Chemistry, 386, 1025e1041.
Braungart, M., & McDonough, W. (2002). Cradle to cradle: Remaking
the way we make things. North Point Press.
Braungart, M., McDonough, W., & Bollinger, A. (2006). Cradle-to-
cradle design: creating healthy emissions e a strategy for eco-
effective product and system design. Journal of Cleaner
Production, 1e12.
CheckPack. (2013). CheckPack. Retrieved 3 2014, from: https://www.
checkpack.ugent.be/.
Clark, D. A. (2010). Major trends in gravure printed electronics.
California Polytechnic State University.
Das, R., & Harrop, P. (2011). Printed, organic & flexible electronics
forecasts, players & opportunities 2011e2021. IDTechEx.
EC. (2009). EU guidance to the commission regulation on active and
intelligent materials and articles intended to come into contact
with food. European Commission.
Esteves da Silva, J. C., & Goncalves, H. M. (2011). Analytical and
bioanalytical applications of carbon dots. TrAC Trends in
Analytical Chemistry, 8(3), 1327e1336.
Eulises, U. (2013). Fullerenes and their applications in science and
technology. Introduction to nanotechnology. Florida International
University.
Gardner, J. W., & Bartlett, P. N. (1993). A brief history of electronic
noses. Sensors and Actuators B, 18, 211e220.
Grynko, D., Burlachenko, J., Kukla, O., Kruglenko, I., & Belyaev, O.
(2009). Fullerene and fullerene-aluminum nanostructured films as
sensitive layers for gas sensors. Semiconductor Physics, Quantum
Electronics & Optoelectronics, 12, 287e289.
Heising, J. K., Dekker, M., Bartels, P. V., & Van Boekel, M. A. (2014).
Monitoring the quality of perishable foods: opportunities for
intelligent packaging. Critical Reviews in Food Science and
Nutrition, 54(5), 645e654.
Ilie-Zudor, E., Kem
eny, Z., Egri, P., & Monostori, L. (2006). In The
modern information technology in the innovation processes of the
industrial enterprises-MITIP (pp. 29e36).
Kaur, M., Sandhu, M., Mohan, M., & Sandhu, P. S. (2011). RFID
technology principles, advantages, limitations & its applications.
International Journal of Computer and Electrical Engineering, 3,
1793e8163.
Kerry, J. P., OGrady, M. N., & Hogan, S. A. (2006). Past, current and
potential utilisation of active and intelligent packaging systems for
meat and muscle-based products: a review. Meat Science, 74,
113e130.
Kruss, S., Hilmer, A. J., Zhang, J., Reuel, N. F., Mu, B., & Strano, M. S.
(2013). Carbon nanotubes as optical biomedical sensors.
Advanced Drug Delivery Reviews, 65, 1933e1950.
Kuswandi, B., Maryska, C., Jayus, Abdullah, A., & Heng, L. Y.
(2013). Real time on-package freshness indicator for guavas
packaging. Journal of Food Measurement and Characterization,
7, 29e39.
Kuswandi, B., Wicaksono, Y., Abdullah, A., Heng, L. Y., & Ahmad, M.
(2011). Smart packaging: sensors for monitoring of food quality.
Sensing & Instrumentation Food Quality and Safety, 5, 137e146.
Llobet, E. (2013). Gas sensors using carbon nanomaterials: a review.
Sensors and Actuators B, 179, 32e45.
Malaysia, U. T. (2013). OpenCourseWare. Retrieved October 2013,
from Immobilization of biological sensing elements: http://ocw.
utm.my/file.php/74/CHAPTER_3_OCW.pdf.
McCartney, M. (2006). Integrating RFID with plastic products and
packaging in the retail supply chain. QLM Consulting.
McDonough, W., Braungart, M., Anastas, P., & Zimmerman, J. (2003).
Cradle-to-cradle design and the principles of green design e
toward new perspectives and practices for engineering and design.
Environmental Science and Technology, 434Ae441A.
62 M. Vanderroost et al. / Trends in Food Science & Technology 39 (2014) 47e62
McFarlane, D., & Sheffi, Y. (2003). The impact of automatic
identification on supply chain operations. International Journal of
Logistics Management, 14, 1e17.
Miltz, J., Passy, N., & Mannheim, C. (1995). Trends and applications of
active packaging systems. In M. Jagerstad, & P. Ackerman (Eds.),
Food and packaging materialsdChemical interaction, Vol. 162
(pp. 201e210). London, England: The Royal Soc. of Chemistry.
Mirica, K. A., Azzarelli, J. M., Weis, J. G., Schnorr, J. M., &
Swager, T. M. (2013). Rapid prototyping of carbon-based
chemiresistive gas sensors on paper. PNAS, E3265eE3270.
Monereo, O., Boix, M., Claramunt, S., Prades, J. D., Cornet, A.,
Cirera, A., et al. (2011). Advanced performances in gas sensors:
stretchable, flexible, wireless, wearable. Procedia Engineering, 25,
1425e1428.
Mottola, F. (2007). Metal-cladding Leaky Waveguides for sensing
applications. Seconda Universita Degli Studi Di Napoli.
Nitkowski, A., Chen, L., & Lipson, M. (2008). Cavity-enhanced on-
chip absorption spectroscopy using microring resonators. Optical
Society of America, 16.
Ortega-Rivas, E. (2012). Non-thermal food Engineering operations.
Springer New York.
Ouellette, J. (1999). Electronic noses Sniff out new markets. The
Industrial Physicist, 5-1, 26.
Pavelkov
a, A. (2012). Intelligent packaging as device for monitoring of
risk factors in food. Journal of Microbiology, Biotechnology and
Food Sciences, 2, 282e292.
Pavelkov
a, A. (2013). Time temperature indicators as devices
intelligent packaging. Acta Universitatis Agriculturae et
Silviculturae Mendelianae Brunensis, 56, 245e252.
Pilollo, R., Monaci, L., & Visconti, A. (2013). Advances in biosensor
development based on integrating nanotechnology and applied to
food-allergen management. Trends in Analytical Chemistry, 47,
12e26.
Potyrailo, R. A., Nagraj, N., Tang, Z., Mondello, F. J., Surman, C., &
Morris, W. (2012). Battery-free radio frequency identification
(RFID) sensors for food quality and safety. Journal of Agricultural
and Food Chemistry, 60, 8535e8543.
Puyol, M., del Valle, M., Garc
es, I., Villuendas, F., Dom
nguez, C., &
Alonso, J. (1999). Integrated waveguide absorbance optode for
chemical sensing. Analytical Chemistry, 5037e5044.
Reyes De Corcuera, J. I., & Cavalieri, R. P. (2003). Biosensors. In
D. Heldman (Ed.), Encyclopedia of agricultural, food, and
biological engineering (pp. 119e123). Marcel Dekker.
Sample, A. P., Yeager, D. J., Powledge, P. S., Mamishev, A. V., &
Smith, J. R. (2008). Design of an RFID-based battery-free
programmable sensing platform. IEEE Transactions on
Instrumentation and Measurements, 57-11, 2608e2615.
Sarley, A. (2011). Implications of thermochromic ink. Graphic
Communication Department; California Polytechnic State
University.
SCDigest Editorial Staff. (2009). Looking back at the Wal-Mart RFID
time line. Supply Chain Digest.
Schedin, K., Geim, A. K., Morozov, S. V., Hill, E. W., & Blake, P.
(2007). Detection of individual gas molecules adsorbed on
graphene. Nature Materials, 6, 652e655.
Scheelite Technologies, L. (2011). Flex alert. Retrieved October 28,
2013, from: http://www2.flex-alert.com/flexalert/wp-content/
uploads/2011/11/Flex-Alert-Flyer-v2.pdf.
Smits, E., Schram, J., Nagelkerke, M., Kusters, R., van Heck, G., van
Acht, V., et al. (2012). Development of printed RFID sensor tags for
smart food packaging. In IMCS 2012 e the 14th International
Meeting on Chemical Sensors. Nuremberg: ICMS.
Stouthuysen, P., & le Roy, D. (2010). Cradle to cradle: theoretical
framework. C2C Network.
Takhistov, P. (2009). Biosensor technology for food packaging
applications. In K. L. Yam (Ed.), The Wiley encyclopedia of
packaging technology (pp. 121). John Wiley & Sons, Inc.
Turner, A. F. (2012). Biosensors: sense and sensibility. Chemical
Society Reviews, 42, 3184e3196.
Verfaille, H. A., & Bidwell, R. (2000). Measuring eco-efficiency: A
guide to company performance. World Business Council for
Sustainable Development.
Vu, C., & Won, K. (2013). Novel water-resistant UV-activated oxygen
indicator for intelligent food packaging. Food Chemistry, 140,
52e56.
Yam, K. L. (2012). Intelligent packaging to enhance food safety and
quality. In K. L. Yam, & D. S. Lee (Eds.), Emerging food packaging
technologies: Principles and Practice (Chapter 8). Woodhead
Publisher, 2012.
Yam, K. L., Takhistov, P. T., & Miltz, J. (2005). Intelligent packaging:
concepts and applications. Journal of Food Science, 70-1,
R1eR10.
Yebo, N. A., Sree, S. P., Levrau, E., Detavernier, C., Hens, Z.,
Martens, J. A., et al. (2012). Selective and reversible ammonia gas
detection with nanoporous film functionalized silicon photonic
micro-ring resonator. Optics Express, 20, 11855e11862.