Food and Nutrition Sciences, 2013, 4, 113-123
http://dx.doi.org/10.4236/fns.2013.47A014 Published Online July 2013 (http://www.scirp.org/journal/fns)
Milk Spoilage: Methods and Practices of Detecting Milk
Quality
Michael Lu, Yvonne Shiau, Jacklyn Wong, Raishay Lin, Hannah Kravis, Thomas Blackmon,
Tanya Pakzad, Tiffany Jen, Amy Cheng, Jonathan Chang, Erin Ong, Nima Sarfaraz, Nam Sun Wang*
Department of Chemical & Biomolecular Engineering, University of Maryland, College Park, USA.
Received March 28th, 2013; revised April 28th, 2013; accepted May 5th, 2013
Copyright © 2013 Michael Lu et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ABSTRACT
Milk spoilage is an indefinite term and difficult to measure with accuracy. This uncertainty can cause suffering for both
milk manufacturers and consumers. Consumers who have been misled by ambiguous expiration dates on milk cartons
waste resources by disposing of unspoiled milk or experience discomfort from drinking spoiled milk. Consumers are
often unwilling to purchase products close to their inaccurate expiration dates. This consumer behavior has a negative
financial impact on milk producers. Inaccurate milk spoilage detection methods also force milk producers to use overly
conservative expiration dates in an effort to avoid the legal and economic consequences of consumers experiencing ill-
ness from drinking spoiled milk. Over the last decade, new methods have been researched with the purpose of develop-
ing more accurate and efficient means of detecting milk spoilage. These methods include indicators based on pH bacte-
ria counts and gas-sensor arrays. This article explores various methods of spoilage detection designed to prevent such
consequences. The respective level of effectiveness of each method is discussed, as well as several further approaches
to contain freshness regardless of detection.
Keywords: Milk Spoilage; Detection; pH; pH Detection; Methylene Blue Reduction; Amperometric Sensor;
Magnetoelastic; Gas Sensor Array; Infrared Spectroscopy; Lipid/Fat Count
1. Introduction
Consumers currently determine milk spoilage by check-
ing the “sell by” and “best if used by” dates on milk car-
tons provided by suppliers. These dates are simple esti-
mates of milk shelf life and are often inaccurate due to
the variable processing, shipping, and storage conditions
of the milk [1]. Retailers and consumers discard billions
of pounds of unspoiled milk each year while relying on
inaccurate printed expiration dates. Conversely, milk
may spoil before the printed expiration, and spoiled milk
can lead to food poisoning if consumed [2]. Another is-
sue demonstrating the urgency of developing more accu-
rate milk packaging is that unless there are local restric-
tions for dairy products, the manufacturer determines the
date; if an established regulation does not exist, manu-
facturers can legally sell the expired product past the
posted date. The cities or processors that do have regula-
tions have different sets of rules. For example, until 2010,
milk could only be legally sold in New York City up to
96 hours after 6:00 AM on the day after pasteurization [3]
[4]. Dairylea Cooperative, Inc., the biggest processor in
the region, allows milk to be sold 10 to 12 days after
pasteurization [5].
Researchers have recently begun investigating appli-
cations of current technologies for the detection of milk
spoilage. While the meat, fish, and fruit industries have
continually advanced new methods of packaging, pack-
aging innovation in the milk packaging industry has re-
mained stagnant [6]. Although a handful of packaging
companies have proposed ideas for updated milk pack-
aging designs, none of the proposed alternatives have
succeeded in the market. Food industries and consumers
have shown a trend of increasing interest in environmen-
tally friendly, health-conscious intelligent packaging, and
these market trends indicate the need to invest heavily in
these novel developments [7]. Innovative milk packaging
will further revolutionize the packaging industry and
transform the way consumers think about packaging.
The aims of this article are to review the current state
of the intelligent food packaging industry and present an
*Corresponding author.

Milk Spoilage: Methods and Practices of Detecting Milk Quality 114
update on the considerations that intelligent food pack-
aging developers may contemplate for further advance-
ment in this field.
2. Current Methods of Milk Spoilage
Detection
2.1. Utilizing pH Indicators as a Measure of
Spoilage
Bacteria growth varies from one species of bacteria to
another. While one bacteria species may prosper under
certain conditions, another species may weaken. These
conditions are interdependent and include nutrient avail-
ability, moisture, oxygen levels and the level of other
gases, the presence of inhibitors, temperature, and pH
[8].
The pH of unspoiled milk is approximately 6.7, a level
at which many forms of bacteria thrive [9]. At lower pH
levels of 4.0 - 5.0, lactic acid bacteria can grow and pro-
duce lactic acid. While these organisms inhibit the
growth of many pathogenic bacteria and are also inten-
tionally employed to ferment milk to make other dairy
products such as yogurt and cheese, they can also induce
undesirable spoilage in certain products.
Coliforms, a common form of bacteria, have been an
indicator of the presence of pathogens in assessing the
contamination of water as well as dairy products [10].
Coliforms can cause rapid spoilage in milk because they
ferment lactose with the production of acid and gas, and
they can also degrade milk proteins. Escherichia coli is a
well-known example of a coliform [11]. Studies have
shown that other properties of milk also promote bacteria
growth, such as the high availability of moisture and
dissolved oxygen which supports both aerobic and facul-
tative anaerobic microorganisms [12]. Temperature is
frequently controlled to limit bacteria growth. Extreme
heat is lethal to many organisms, such as coliforms,
which explains the process of milk pasteurization (63˚C
for 30 minutes). Two types of bacteria exist in pasteur-
ized milk: thermoduric bacteria, which are capable of
surviving the extreme heat during pasteurization, and
bacteria that originate from unsanitary conditions post-
pasteurization [13]. Psychrotrophs comprise the largest
percentage of bacteria in milk and cause spoilage in re-
frigerator temperatures at or below 7˚C [14].
Acidity increases as milk spoils; thus, acidity can be
quantified to measure milk quality. Acidity in dairy prod-
ucts can be expressed in two ways: 1) titratable acidity,
which shows total acidity but not acid strength; and 2)
hydrogen ion concentration or pH, which indicates acid
strength. The natural acidity of milk is 0.16% - 0.18%,
and samples with higher figures indicate developed acid-
ity [15]. At normal levels of pH, the main protein in milk,
casein, remains evenly dispersed. At lower levels of pH
below 4.6, the protein can no longer remain in solution,
so it coagulates due to acid generated from fermentation.
Two studies confirm the link between pH change in
milk and spoilage: Fromm and Boor (2004) researched
the attributes of pasteurized fluid milk (2% High Tem-
perature/Short Time, HTST milk) during its shelf life.
Milk samples were randomly collected from three fluid
milk processing plants in the state of New York. A group
of 13 panelists evaluated 2% HTST processed fluid milk
products based on a quantitative descriptive analysis.
They tasted and scored the perceived intensity of the
aroma, taste, and aftertaste of milk samples varying in
degree of freshness using a numeric scale ranging from 0
to 15 and the descriptive terms listed in Table 1.
The free fatty acid (FFA) content of the samples sig-
nificantly increased throughout shelf life. Though not
significantly different between day one and day seven,
the FFA content drastically increased between days four-
teen to seventeen due to milkfat lipolysis. The higher the
FFA, the more likely sensory panelists were able to de-
tect lipolyzed or rancid off-flavors in 2% fat milk [16].
Casein levels in all milk samples also decreased ap-
proximately 2% at a relatively rapid rate following sev-
Table 1. Descriptors of attributes of pasteurized fluid milk
by Fromm and Boor (2004).
Descriptors
CategoryAttribute Reference for intensity of attribute
Cheese Colby cheese
Cooked Freshly pasteurized milk (79˚C/18s)
Cream Heavy cream
Hay/grain Hay
Sulfur Boiled egg
Sour/fermented Sauerkraut
Aroma
Putrid Limburger cheese
Baby formulaBaby formula
Butter Butter
Cooked Freshly pasteurized milk (79˚C/18s)
Flat 2% fat milk with 3% added water
Nutty Peanuts
Rancid Strong provolone cheese
Taste
Sweet 1:1 100% lactose free Lactaid
and 2% fat milk
Cardboard Cardboard box
Sweet 1:1 100% lactose free Lactaid
and 2% fat milk
Sour Cultured buttermilk
Metallic FeSO4 (0.445 g/L)
Drying Unsweetened tea
Aftertaste
Lingering -

Milk Spoilage: Methods and Practices of Detecting Milk Quality 115
enteen days of refrigeration. This is associated with off-
flavors in fluid milk, particularly bitterness.
This study concludes that each processing plant has
different microflora species and needs to have plant-
specific strategies to identify and reduce sources of con-
tamination. However, these species, while different, all
cause milk to decrease in pH. Increases in FFA and drops
in casein levels correlate with a decrease in pH. This
suggests that pH can serve as a measurement not only of
milk spoilage but also of milk edibility, since panelists
determined FFA and casein levels affect rancidity and off
flavors in fluid milk [17]. This specific study, however,
does not establish a lactic acid level and corresponding
pH at which milk remains drinkable. Though pH meters
are available, modern versions are inconvenient and cum-
bersome for individual consumers.
Ostlie, Helland, and Narvhus (2003) conducted a study
to analyze the amount of metabolic products produced by
five specific probiotic strains in ultra-high temperature
(UHT) treated milk [18]. The study used a pH meter with
a combined glass electrode and temperature probe to
measure pH during fermentation. Volatile compounds
were analyzed with headspace gas chromatography and
organic acids were analyzed with high pressure liquid
chromatography. Quantitative analysis of carbon dioxide
production was determined with an infrared CO2 gas
analyzer. Below, Table 2 shows the results of survival
and storage stability:
Preliminary studies showed that the growth varied
considerably with the concentration of the added sup-
plements. After 6 - 16 hours of incubation, all strains at-
tained viable cell numbers above 8.7 - 9.18 log CFU/mL.
Depending on the probiotic strain used, the pH of the
ultra-high temperature milk decreased from 6.7 initially
to 3.9 - 4.4 after 24 hours of incubation.
A disadvantage of the study was that in fortified milk,
the various probiotic strains possessed different meta-
bolic profiles, which affected the sensory quality of
products containing these different organisms. The in-
crease in strain growth (cell numbers), led to an increase
in the amount of lactic acid produced. This increase in
lactic acid in turn led to a drop in pH. Whereas the earlier
methodology correlated levels of pH with amounts of
protein, this study correlated pH directly with the number
of bacteria cells.
These studies conclude pH is quantifiable and can
measure spoilage in milk. However, there must be further
research to address the shift in metabolism of probiotic
bacteria in response to environmental changes, as well as
the effect of different milk treatment regimes on the me-
tabolism of probiotic bacteria in milk. In addition, the pH
range where milk meets the definition of “spoiled” does
not coincide with the colloquial definition of spoilage. As
in the earlier study, most consumers consider milk to be
inedible at those pH levels. For the future development
of pH as an indicator of milk quality, a more accurate pH
range must be established to define the point at which
milk is no longer drinkable. This goal can be accom-
plished by combining Fromm and Boor’s methodology
involving a group of panelists to determine the level of
consumable spoilage with Ostlie, Helland, and Narvhus’
study of pH values and acidic byproducts.
There are currently a few devices that can determine
the acidity levels of milk. These are typically used as a
means of quality control by manufacturers, rather than by
end consumers. Many pH electrodes/meters and titra-
tors/meters quickly and accurately measure pH of dairy
products. A prototype, the Milkmaid smart jug, is a new
product that detects when milk starts to spoil with a pH
sensor in the base. The jug informs users of spoilage via
a change in the color of the jug’s light-emitting diode
lights. This product is not yet sold and the price has not
yet been determined. A disadvantage of this design is
that consumers must pour milk into a separate container,
because the pH sensor is not incorporated into the plastic
container milk is sold in. Thus, while the Milkmaid jug
shows that detecting spoilage with pH is applicable to
commercial products, it remains to be seen if this par-
ticular design becomes commercially successful.
Table 2. Viable pH counts in UHT milk (log·cfu·ml−1)2.
Time of freezing
0 hour 1 day 1 month 6 months
Sucrose (%)
Strain 0 0 5 10 0 5 10 0 5 10
Lb. acidophilus La5 8.78 8.63 8.70 8.76 8.79 8.84 8.81 8.70 8.79 8.80
Lb. acidophilus 1748 10.11 - - - 10.18 - - - - -
Lb. rhamnosus GG 9.08 9.04 9.06 9.09 9.15 9.05 9.08 9.17 9.16 9.25
Lb. reuteri SD 2112 9.17 9.05 9.08 9.02 9.15 9.05 9.08 9.17 9.16 9.25
B. animalis BB12 8.86 8.84 8.83 8.80 8.93 8.93 8.87 8.75 8.71 8.76

Milk Spoilage: Methods and Practices of Detecting Milk Quality 116
2.2. Electrical Methods for the Detection of
Bacteria
Some traditional methods of detection involve bacterial
enumeration, in which spoilage is detected when in-
creased metabolism caused by multiplying bacteria ren-
ders a colored solution colorless. The methylene blue
reduction test is such an example; however, known flaws
of this test include time-consuming and redundant pro-
cedures, as well as an inability to discriminate between
bacterial types [19]. Lee et al. (2009) sought to improve
upon the methylene blue reduction method while main-
taining its advantages by supplementing it with an am-
perometric sensor. An amperometric sensor, composed
of a circuit with a potentiostat and a pair of electrodes,
measures current change [20]. Amperometric sensors are
small and inexpensive and have been tested in a variety
of media to detect changes in bacteria such as E. coli.
Lee et al. inoculated with milk E. coli and Ent. aero-
genes, two types of coliforms that indicate the sanitary
condition. A third sample contained milk and methylene
blue. Methylene blue is blue until the metabolic activity
of bacteria causes it to lose color. Consequently, the bac-
terial metabolism of the E. coli caused the reduction of
methylene blue in the three samples and also resulted in a
current change. Any current change of more than 0.05
μA was detected with the amperometric sensor and re-
corded. The study tracked detection time and provided an
estimate of the approximate number of microorganisms
initially in the sample. Results had R2 of 0.9192, cor-
roborating high accuracy in an inverse linear relationship
between the log of the bacterial concentration against the
detection time. The increase of microbial organisms ex-
ponentially related to the time from inoculation to the
initial small change in current.
Results were favorable. Advantages to this method in-
clude a detection time 0.5 - 2 hours shorter than that ob-
tained with the methylene blue reduction method and a
very broad detection range of 102 - 104 CFU/mL. Fur-
thermore, whereas the methylene blue reduction method
required constant supervision and sampling at a 30-min-
ute interval, the amperometric sensor could independ-
ently record the data. The latter procedure was relatively
simple and inexpensive; accuracy was also a non-issue.
However, this method cannot discriminate between
viable and non-viable cells. Furthermore, type of bacteria
detection was lacking. The amperometric sensor could
only detect E. coli and Ent. aerogenes coliforms; when
other bacteria such as B. subtilis, Lactobacillus sp., Sac-
charomyces sp., and Staph. aureus were tested upon,
they produced a negligible current change.
2.3. Wireless Detection and Monitoring of Milk
Spoilage
Application of remote-query technology to detect milk
spoilage is an emerging field of experimentation. The
remote-query magnetoelastic sensor platform is a free
standing, ribbon-like magnetoelastic thick-film coupled
with a chemical or biochemical sensing layer such as an
enzyme that vibrates at a characteristic resonance fre-
quency. A pickup coil is then used to remotely detect the
magnetic field generated by the mechanical oscillations.
Magnetoelastic sensors have already been developed and
tried in a number of different types of analyses; past re-
search has used it for the analysis of glucose concentra-
tion, blood clotting, and detection of Escherichia coli as
well as Salmonella typhimurium [21-24]. However, it has
not directly been applied to the detection of milk spoil-
age.
Huang et al. (2008) tested a remote-query magnetoe-
lastic sensing platform to quantify the bacterial count of
Staphylococcus aureus ssp. anaerobius (S. aureus) in
milk. S. aureus is a bacterium that resides in milk and
multiplies as milk spoils; infection with S. aureus can
result in such human diseases as toxic shock syndrome,
endocarditis, and septicemia [25]. Culture media of S.
aureus were prepared, and resonance characteristics were
measured with a magnetoelastic sensor fabricated from
Metglas alloy ribbon. Because the sensor responds to
mass loading due to bacterial adhesion and changes in
solution viscosity, Huang et al. increased the viscosity of
one of the culture media—one a nutrient broth and one a
tryptic soy broth—through different trials; ultimately, the
sensor showed a higher sensitivity in the milk than in the
culture medium because of the higher viscosity of milk.
Results concluded that the sensor platform is feasible for
use in the remote detection of spoiled milk samples.
Remote-query detection is, overall, an effective me-
thod. Typically, material costs are low; disposable mag-
netoelastic sensors, for example, can be fabricated from
strips costing $300 per kilometer, while the combination
of an amperometric sensor with methylene blue also in-
volves low costs. However, deficiencies in this method
continue to surface. As bacteria detection becomes com-
plex, as with the ATP (adenosine triphosphate) biolumi-
nescence method and the PCR (polymerase chain reac-
tion) method, speed becomes an issue [26,27]. The sim-
ple methylene blue reduction method was ponderous as
bacteria detection requires constant supervision.
The two methods presented attempts to improve upon
these inefficiencies. Not only does Lee et al.’s experi-
mentation with the amperometric sensor provide a broad
detection range between 102 - 104 CFU/mL, electro-
chemical techniques are easier to apply and have lower
costs. Lee’s method solved the problem of speed and,
more importantly, provided a user-friendly procedure.
2.4. Gas-Sensor Arrays
Haugen, Knut, Langsrud, and Bredholt (2006) attempted

Milk Spoilage: Methods and Practices of Detecting Milk Quality 117
to predict the shelf-life of milk with gas-sensor arrays. In
their experiment, a commercial solid-state gas-sensor
array system monitored the growth of disinfectant-resis-
tant bacteria in milk known to cause spoilage, namely
Serratia marcescens, Serratia proteamacufans, and Pseu-
domonas putida.
The Haugen study identified the quality and amount of
major volatile microbial metabolites using gas chroma-
tography and mass spectroscopy, and also analyzed and
correlated microbiological data through statistical tests.
Haugen et al. patterned and fingerprinted complicated
odors with gas sensor arrays in order to determine the
correlation between secondary volatile metabolites from
the milk and the microorganisms that produced these
compounds. This method allows the determination of
organisms responsible for spoilage, quality of the milk,
and length of shelf life based on objective criteria [28].
Twenty-six Gram-negative bacteria were selected and
identified based on cellular fatty acid analysis, and three
spoilage-inducing isolates (Cedcea sp., Serratia sp., and
Pseudomonas sp.) that proliferate in milk were chosen
for experimentation.
Aside from the CO2 signal, which stayed constant at
baseline level, the sensor signals decreased during the
first 6 - 7 hours because background volatile compounds
from the medium were extracted dynamically from the
cultures during each measurement. The production of
volatile secondary bacterial metabolites started to exceed
the background volatiles and increased significantly
around 7 - 8 hours (excluding the pure Pseudomonas
culture). This increase in signal coincided with the onset
of the exponential growth phase. At 18 - 20 hours, a peak
in CO2 production corresponded to a plateau in S. mar-
cescens N9, S. proteamaculans O6A and the mixed cul-
tures. The sensor signals from the pure cultures showed
significant correlation with the cell counts.
This gas-sensor array system successfully detected
major volatile metabolites produced by bacteria during
growth. A significant majority of the contributions to the
sensor response signals were ascribed to the major vola-
tile compounds except for acetate. Haugen et al. investi-
gated the possibility of developing a system that detects
and monitors growth of undesirable bacteria. The high
correlation between the sensor readings and the cell
counts of the pure cultures suggests that gas-sensor mea-
surements can predict bacterial cell numbers in pure cul-
tures. Therefore, a design of specific sensors can be
adapted to follow the development of spoilage organisms
in specific food products.
Secondary compounds identified by gas chromatogra-
phy and mass spectroscopy comprised of fermentation
products produced by Serratia, Enterobacter, and Er-
winia. Both Enterobacteriaceae and Ps. putida produced
spoilage off-odors in pasteurized milk, but the Ps. pu tida
culture contributed a far lower production of volatiles
and CO2 than the Serratia strains at 25˚C. At less than
10˚C, Ps. putida would have been the major spoilage
organism in milk [29]. Therefore, sensors must be se-
lected and developed to provide indicators of typical tar-
get organisms in different food products under specified
temperature, pH, and water activity conditions.
However, there are constraints to detecting bacterial
strains in complex cultures based on temporal gas sensor
response patterns from single strains. In particular, Ps.
Putida produced secondary volatile metabolites and only
contributed to less than 0.1% of the temporal response
pattern, even though the Ps. putida reached a significant
concentration in the mixed culture. Therefore, for the
applied experimental conditions, the gas sensors were not
sufficiently sensitive to detect Ps. putida in the mixed
culture, because the vapor phase was masked by the
volatile metabolites generated by S. marcescens N9. The
sensors also have a certain extent of cross-sensitivity, so
it is important to select sensors with high sensitivity for
detection of strain specific metabolites, especially when
the spoilage bacteria constitute a lesser percentage of the
total flora.
Gas sensors can be designed to detect the characteris-
tic metabolites of specific bacteria in specific food prod-
ucts and can offer rapid, accurate determinations of shelf
life. To detect correctly target strains of bacteria, it is ne-
cessary for sufficient strain specific volatile compounds
to be present. Thus, a drawback of gas sensor technology
is its great potential for strains where the strain specific
metabolites represent the major volatile compounds in
the vapor phase. When bacteria do not produce many
volatile compounds, more sensitive sensors are needed to
detect their characteristic metabolites.
2.5. Infrared Spectroscopy as Spoilage Indicator
Spectroscopy is a nondestructive technique, where spec-
tral features provide biochemical information regarding
the molecular interactions between, and the composition
and structure of, different cells and tissues. This method
was first widely applied in the food industry to detect
spoilage in beef, rainbow trout fillets, and other meat
products [30,31]. However, this method had not been
experimented on milk until Al-Qadiri, M. Lin, Al-Holy et
al. in 2008. To do so, they evaluated visible and short
wavelength near-infrared diffuse reflectance spectros-
copy (SW-NIR) as a technique to detect milk spoilage in
pasteurized skim milk. They wanted to see the feasibility
of applying visible and SW-NIR spectroscopy to monitor
spoilage of pasteurized skim milk in industrial settings.
In doing so, Al-Qadiri et al. first took the total aerobic
plate count and pH measurements. They then examined
the milk samples at 22˚C to correct for spectral changes
that could result from temperature differences during

Milk Spoilage: Methods and Practices of Detecting Milk Quality 118
spectral collection. The mean pH measurement for con-
trol milk samples was 6.66, and they found no obvious
pH decrease for milk samples stored at 6˚C after 30
hours of storage. In experimental samples, the visible and
SW-NIR diffuse spectroscopy detected the formation of
metabolic byproducts from proteolysis and lipolysis
caused by bacterial cell growth, which led to a reduction
in pH [32]. This method was effective, but costly. Fur-
ther work will be needed to identify which biochemical
changes in spoilage microorganismss correlate with spe-
cific SW-NIR spectral features.
Nicolaou et al. (2012) attempted to take infrared spec-
troscopy further with matrix-assisted laser desorption/
ionization time-of-flight mass spectrometry (MALDI-
TOF-MS). MALDI-TOF-MS has already been used in
protein and peptide identification and quantification;
however, Nicolaou wanted to see if it was useful for mi-
crobial spoilage assessment because techniques for iden-
tifying and quantifying spoilage bacteria in pasteurized
milk are time-consuming. Their methodology included
incubating milk samples and raw pork meat samples at
15˚C and at room temperature, and then analyzing them
with MALDI-TOF-MS at a rate of 4-minute intervals.
MALDI-TOF-MS has many advantages, particularly
in terms of sensitivity, accuracy, and speed. Spectrum
can be generated within minutes following sample prepa-
ration. Its most comparable technology is fourier trans-
form infrared (FT-IR) spectroscopy; however, MS allows
more equivocal identification of important proteins while
FT-IR spectroscopy does not, or does so at best only em-
pirically through peak assignments [33]. Furthermore,
MALDI-TOF-MS has minimal sample preparation, which
contributes to the rapid speed of data collection. Typical
sample speed is 4 minutes per sample, which is consid-
erably faster than classical microbiological plating ap-
proaches that can take up to 2 days. Drawbacks, how-
ever, include the limited use of infrared spectroscopy in
the field. The technology is perceived as a tool for as-
sessing protein qualitatively rather than for measuring
microbial bacterial count quantitatively. Familiarization
of MALDI-TOF-MS can help change perceptions and
lead to use of this technology in the dairy industry.
However, the technical difficulty of this method renders
it unsuitable for consumer use.
2.6. Protein or Fat Count Detection
Several studies correlate bacteria that are known to be
involved in the spoilage of pasteurized milk with varia-
tions in the levels of lipids and proteins that are present
in milk. Yagoub, Bellow, and El Zubeir (2008) con-
cluded that the lowest percentages of lipid and protein in
milk occur when Pseudomonas levels are at their highest.
High Pseudomonas levels cause high levels of proteolytic
activity in all food systems. Lipases and proteases break
down lipids and proteins, which lowers lipid and protein
levels. The by-products of these hydrolysis reactions in-
crease milk acidity and directly correlate to milk spoilage.
Additionally, the bacteria Pseudomonas aeruginosa chan-
ge the constituents of milk, thereby spoiling it [34].
There are many common industry practices that in-
corporate this method into testing milk properties. Pro-
tein levels are typically determined by the standard
Kjeldahl method or the more favorable Dumas method
[35]. The Kjeldahl method measures nitrogen levels by
using a fitting titration technique. There are three steps
involved: digestion, neutralization, and titration. The pro-
tein content is calculated from the nitrogen concentration
in the milk. This standard method does not directly meas-
ure protein content and thus needs a conversion factor
(which varies among different proteins) to convert meas-
ured nitrogen concentration to a protein concentration.
This conversion can lead to inaccuracies. Another disad-
vantage is the amount of time (1 - 2 hours) required to
perform this test.
The Dumas method, a similar but enhanced version of
Kjeldahl method, is an automated instrumental technique
that combusts in the presence of oxygen a sample of
known mass in a high temperature chamber. By-products
CO2 and H2O are filtered out, leaving only N2 or nitrogen
content to be read by a thermal conductivity detector.
This test is can be done in fewer than 4 minutes. Like the
Kjeldahl method, the Dumas method also needs to con-
vert nitrogen content to protein content, as it does not
truly measure the protein. The method also carries high
initial costs and can make it difficult to obtain a repre-
sentative sample. Alternate methods of protein counts are
time consuming and require extensive sample prepara-
tion before analysis.
Lipid or fat content is primarily determined by Gerber
method in Europe, or the very similar Babcock method in
the United States. The Gerber method requires adding
dairy product into a butyrometer and adding concentrated
sulfuric acid and amyl alcohol to dissolve the non-fat
milk solids [36]. The mixture is centrifuged for a set time
at 1100 rpm and placed in a water bath to standardize the
samples before reading the fat content off the calibrated
scale of the butyrometer. The method is fast, low cost,
and suitable for high volumes of sample. However, dis-
advantages include not being able to automatically de-
termine levels and the risk of handling highly concen-
trated sulfuric acid. In addition, reading the butyrometer
takes acquired skill, so the method cannot be used by
average consumers.
Sorhaug and Stepaniak (1998) showed that psychotro-
pic Bacillus spp. are resistant to heat treatment and se-
cretes extracellular proteinases, lipases, and phosopli-
pases. Other bacteria that survive the pasteurization proc-
ess also produce these proteins. Additionally, the quality
of dairy products can decrease due to heat-resistant en-

Milk Spoilage: Methods and Practices of Detecting Milk Quality 119
zymes that are secreted by psychrotrophs in raw milk
before heat treatment, or produced by psychrotrophs grow-
ing during the cold refrigeration of dairy products [37].
Prior studies also show that up to 20% of all psychrotro-
phic bacteria isolated from raw milk possess proteolytic
and lipolytic enzymatic activities [38]. Sorhaug and Ste-
paniak determined the average psychrotroph count for
different dairy types at the time of spoilage for Pasteur-
ized milk to be 6 - 7.5 CFU/mL, but they did not record
the protein level. For this method to be used as a spoilage
detector, more testing is needed to determine the equiva-
lent protein level at the psychrotroph count of 6 - 7.5
CFU/mL and the precise correlation between protein
count and spoilage.
Average consumers cannot perform a protein or fat
count due to the method’s required know-how, needed
time, and financial costs. Thus, it is impractical to incor-
porate this method into commercially sold products for
consumers, even in light of the fact that levels of post-
pasteurization contamination correlate extremely well
with shelf life, and protein and fat count is arguably the
most accurate indicator of expected shelf life of pasteur-
ized milks for milk processors. Yagoub, Bellow, and El
Zubeir (2008) also determined that the byproducts of
lipid and protein breakdown induce increased acidity.
3. Comparison of Milk Spoilage Detection
Methods
Accuracy, range, usability, speed, and cost are of par-
ticular importance when choosing a milk spoilage detec-
tion method because these factors impact the feasibility
and marketability of packaging products. Table 3 com-
pares available milk spoilage detection methods in terms
of those five characteristics:
4. Current Real-World Applications of Milk
Spoilage Prevention
While the milk packaging industry has lagged woefully
behind in milk spoilage detection, there has been a recent
push towards innovation in regards to spoilage preven-
tion. These efforts do not eliminate the need for future
research into milk spoilage detection methods, but could
serve as substitute solutions to the problems previously
discussed in this paper. Detailed below are some exam-
ples of these recent innovations.
4.1. Tetra Pak USA
Tetra Pak is a leading food processing and packaging
solutions company in the world that emphasizes innova-
tion and environmental friendliness in their products. In
the general food industry, they have introduced the Tetra
Recart, which is the first retortable carton package de-
signed for shelf-stable products traditionally filled in
cans, glass jars, or pouches. The Tetra Recart guarantees
freshness for up to 24 months. This product is currently
available for use only with shelf-stable products, but the
company has announced further plans to begin develop-
ment on a similar type of carton for milk which would
guarantee freshness for up to 6 months [39].
4.2. Evergreen Packaging, Inc.
Evergreen Packaging, Inc. specializes in dairy, juice, and
liquid packaging and caters to customers specific needs.
Evergreen Packaging focused their milk carton design on
preventing light and temperature changes since dairy
products are sensitive to these changes and milk reacts
negatively to increases in temperature and light exposure.
Evergreen Packaging claims their fiber-based packaging
keeps light and oxygen out while retaining vitamins and
taste, which helps lock in freshness. Specifically, they
have patented their superior oxygen and moisture-barrier
technology [40].
4.3. TempTime Corporation
The TempTime Corporation is an international manufac-
turer specializing in time-temperature sensitive indicators
for food products. Their product, Fresh-Check, is at-
tached to the outside packaging of temperature-sensitive
food products and pharmaceuticals. The World Health
Organization was one of the first users of time tempera-
ture indicators when they applied this technology to en-
sure the effectiveness of their vaccines in Africa. Since
then, this technology has transformed the administration
of vaccines and has been recently applied to the food
industry by the TempTime Corporation [41]. Approxi-
mately the size of a quarter, the Fresh-Check changes
color when the food product on which it is attached is
exposed to unfavorable temperature and is no longer fit
for consumption. A study conducted by the National
Veterinary and Food Research Institute in Finland deter-
mined that a correlation exists between the color change
of the indicator and the sensory and microbiological
quality of the food product tested [42]. Typically priced
between $0.025 and $0.035 per package, the Fresh-
Check indicators are a more accurate representation of
the quality of the food than the use-by-date, which can-
not account for abuses that may have occurred during the
cold chain process [43].
5. Comparison of Intelligent Packaging
Designs for Milk
Intelligent packaging comes with advantages and disad-
vantages. For milk packaging, issues of freshness and
per-carton cost are of particular interest. Table 4 evalu-
ates three intelligent milk packaging designs.

Milk Spoilage: Methods and Practices of Detecting Milk Quality 120
Table 3. Comparison of the relevant characteristics of milk spoilage detection methods.
Characteristics
Methods Accuracy Range Usability Speed Cost
pH pH levels easily fluctuate;
difficult to reach exact reading
Broad and easily
adaptable
Easy to use; can
incorporate into
individual containers
Instantaneous $600 - $1000
Methlyne blue
reduction with
amperometric sensor
Accurate, but cannot
discriminate between
bacteria types
Broad; 102 - 104
CFU/ml
Easy to use;
uncomplicated
1 - 2 hours; requires little
supervision $1500 - $5000
Magnetoelastic
Only experimented on
S. aureus, not on other
types of bacteria in milk
Broad and easily
adaptable
Easy to use;
uncomplicated 18 hours
$300 per km of
Metglas alloy ribbon;
material costs are low
Gas-sensor array Cannot detect low
concentrations of bacteria
Inflexible; vapor
produced bacteria
may mask
Difficult to use;
calibration and further
analysis needed
Instantaneous
measurement with
threshold concentrations
$5000 - $100,0002
Infrared spectroscopy
Very accurate; already
in commercial use
for meat and fish
Broad and
sensitive
Fairly user friendly,
but perceptions prevent
wide use for milk
Generated within 4
minutes
High instrument cost;
low running costs
Protein/fat count
Different correction factors are
needed for different proteins to
account for different amino acid
sequences
Broad and
sensitive
Complicated; requires
technical knowledge
1 - 2 hours for Kjeldahl
method; 4 minutes for
Dumas method
$4500 - $5000
2Cost will likely decrease in the next decade to less than $100 by exploiting conducting polymers.
Table 4. Comparison of currently available or proposed intelligent milk packaging designs.
Advantages and Disadvantages of Intelligent Milk Packaging
Product Description Advantages Disadvantages
Tetra recart by Tetra Pak
Retortable carton package designed for
shelf-stable products traditionally filled in
cans, glass jars, or pouches
 Developing a carton for milk which
would guarantee freshness up to 6
months

 Proven successful in other products
 Better shape for storage than cans
Only applicable to
shelf-stable products at
present
 No indication of spoilage
level
Evergreen Packaging, Inc.
Milk carton designed to minimize light and
temperature changes; fiber-based packaging
that keeps light and oxygen out while
keeping vitamins and taste in
 Patented their barrier technology
 Offer a superior oxygen and
moisture-barrier board
 Relatively expensive on a
per-carton basis
 No indication of spoilage
level
Fresh-check by the
TempTime Corporation
Attached to the outside packaging of
temperature-sensitive food products
and pharmaceuticals
 Reduces food wastage
 Easy to read visual indicator
 Typically priced between. $0.025
and $0.035 per package
 Only based upon temperature
exposure
 Does not actually measure
milk quality levels
6. Conclusions
Existing standards for spoilage detection at milk plants
are obsolete. Often, plants use rudimentary methods such
standard plate count (SPC) to determine and detect bac-
terial concentrations in post-pasteurized milk. These me-
thods are time-consuming and cumbersome, especially
when compared to those used by other food industries.
For example, the meat and fish industries have advanced
to adopting technologies such as infrared spectroscopy to
monitor the quality of the products. Recent research has
illuminated a variety of other methods that can stream-
line the bacterial detection process. Broadly speaking,
these methods include detection based on the pH, current
change, volatile compounds, and lipid and protein levels.
Simplicity of procedure, speed and range of detection,
accuracy of results, and cost of equipment are critical
factors in effectively discerning spoilage in milk. We
considered these criteria to determine the optimal detec-
tion method for spoilage in milk (Table 3).
Comparison of the aforementioned methods shows
that the amperometric sensor combined with the methyl-
ene blue reduction test exhibits high potential for mass
usage at milk processing plants. This method demon-
strates high accuracy and a broad bacterial detection
range (102 - 104 CFU/mL).
Furthermore, this combination of methods has the add-
ed benefit of a user-friendly procedure. Its methodology
is simpler and more easily implemented than that of ei-
ther the gas-sensor array method or spectroscopy meth-

Milk Spoilage: Methods and Practices of Detecting Milk Quality 121
ods. While theoretically the lipid and protein method is
simplistic in nature, actual application generates com-
plexities and safety hazards become a problem.
pH is also a user-friendly method, but its shortcomings
in accuracy would likely be a significant cause for con-
cern among milk plants and manufacturers. Still, pH can
and should be considered for commercial individual use
of spoilage detection because of its low cost and ease-of-
incorporation. Although pH tests have less than perfect
accuracy, they are still reliable indicators of milk shelf-
life. Theoretically, this method can replace expiration
dates as measures of milk spoilage.
At the reasonable price of approximately $3000, the
amperometric sensor compares favorably to the gas-sen-
sor array and the lipid and protein count apparatus. This
price can be considered mid-range, and comprises a one-
time, fixed cost investment for a milk plant.
A major disadvantage of the amperometric sensor is
that it cannot continuously monitor the bacteria growth
of the same sample over time; rather, the sensor can de-
tect only bacteria concentration of the initial sample.
However, this is not a major factor from the perspective
of milk processing plants, which only need to conduct
quality control tests to meet industry standards and then
quickly distribute their pasteurized milk to retailers.
Overall, for simplicity, pH and amperometric sensors
are the optimal methods of milk spoilage detection, but
the success of each method depends on the user’s needs.
For increased reliability, a combination of more than one
method is advised for milk processors. At this moment,
simple indicators continue to evade consumers.
7. Acknowledgements
This study was financially supported by the Gemstone
Program at the University of Maryland, Maryland Tech-
nology Enterprise Institute, and Atlantic Coast Confer-
ence Inter-institutional Academic Collaborative. Addi-
tional thanks to Professor David Bigio of the Department
of Mechanical Engineering.
REFERENCES
[1] J. Kim, D. Twede and J. Lichty, “Consumer Attitudes
About Open Dating Techniques for Packaged Foods and
Over-the-Counter Drugs,” Journal of Food Products Mar-
keting, Vol. 4, No. 1, 1997, pp. 17-30.
doi:10.1300/J038v04n01_03
[2] United States Food and Drug Administration, “Foodborne
Illness-Causing Organisms in the U.S.—What You Need
to Know,” 2008.
http://www.fda.gov/downloads/food/resourcesforyou/con
sumers/ucm187482.pdf
[3] E. B. Hager, “Removing Mysterious Sell-By Date from
Milk Cartons,” The New York Times, 15 June 2010.
http://cityroom.blogs.nytimes.com/2010/06/15/removing-
a-mysterious-sell-by-date-from-milk-cartons/
[4] T. Fleischer, “NYC May Expire Milk Date Rules,” ABC
News, 14 June 2010.
http://abclocal.go.com/wabc/story?section=news/consum
er&id=7497226
[5] F. Fabricant, “Freshness of Milk Products,” The New
York Times, 10 March 1982.
http://www.nytimes.com/1982/03/10/garden/freshness-of-
milk-products.html
[6] I. N. A. Ashie, J. P. Smith, B. K. Simpson and N. F.
Haard, “Spoilage and Shelf-Life Extension of Fresh Fish
and Shellfish,” Critical Reviews in Food Science & Nu-
trition, Vol. 36, No. 1-2, 1996, pp. 87-121.
doi:10.1080/10408399609527720
[7] D. Restuccia, et al., “New EU Regulation Aspects and
Global Market of Active and Intelligent Packaging for
Food Industry Applications,” Food Control, Vol. 21, No.
11, 2010, pp. 1425-1435.
doi:10.1016/j.foodcont.2010.04.028
[8] J. D. Eifert, C. Gennings, W. H. Carter, S. E. Duncan and
C. R. Hackney, “Predictive Model with Improved Statis-
tical Analysis of Interactive Factors Affecting the Growth
of Staphylococcus aureus,” Journal of Food Protection,
Vol. 59, No. 6, 1996, pp. 608-614.
[9] T. R. Klaenhammer, “Bacteriocins of Lactic Acid Bacte-
ria,” Biochimie, Vol. 70, No. 3, 1988, pp. 337-349.
doi:10.1016/0300-9084(88)90206-4
[10] J. L. Kornacki and J. L. Johnson, “Enterobacteriaceae,
Coliforms, and Escherichia coli as Quality and Safety In-
dicators,” Compendium of Methods for the Microbiologi-
cal Examination of Foods, Vol. 4, 2001, pp. 69-82.
[11] United States Department of Agriculture Agricultural
Marketing Service, “Commodity Areas,” 2013.
http://www.ams.usda.gov/AMSv1.0/dairy
[12] R. P. Singh and B. A. Anderson, “The Major Types of
Food Spoilage: An Overview,” In: R. Steele, Ed., Under-
standing and Measuring the Shelf-Life of Food, CRC
Press, Boca Raton, 2004, pp. 3-23.
doi:10.1533/9781855739024.1.3
[13] D. R. Sepulveda, M. M. Góngora-Nieto, J. A. Guerrero
and G. V. Barbosa-Cánovas, “Production of Extended-
Shelf Life Milk by Processing Pasteurized Milk with
Pulsed Electric Fields,” Journal of Food Engineering,
Vol. 67, No. 1, 2005, pp. 81-86.
doi:10.1016/j.jfoodeng.2004.05.056
[14] E. Lopez and R. Lopez, “Demand for Differentiated Milk
Products: Implications for Price Competition,” Agribusi-
ness, Vol. 25, No. 4, 2009, pp. 453-465.
doi:10.1002/agr.20219
[15] Ministry of Agriculture, “Training Programme for Small
Scale Dairy Sector and Dairy Training Institute,” In: Milk
Processing Guide Series 2: Milk Testing and Quality
Control, FAO/TCP/KEN/6611 Project, Naivasha, 2013.
http://www.fao.org/ag/againfo/resources/documents/MPG
uide/mpguide2.htm
[16] H. I. Fromm and K. J. Boor, “Characterization of Pas-
teurized Fluid Milk Shelf-Life Attributes,” Journal of
Food Science, Vol. 69, No. 8, 2013, pp. 207-214.

Milk Spoilage: Methods and Practices of Detecting Milk Quality 122
doi:10.1111/j.1365-2621.2004.tb09889.x
[17] M. V. Santos, Y. Ma, Z. Caplan and D. M. Barbano,
“Sensory Threshold of Off-Flavors Caused by Proteolysis
and Lipolysis in Milk,” Journal of Dairy Science, Vol. 86,
No. 5, 2003, pp. 1601-1607.
doi:10.3168/jds.S0022-0302(03)73745-X
[18] H. M. Ostlie, M. H. Helland, M. H. and J. A. Narvhus,
“Growth and Metabolism of Selected Strains of Probiotic
Bacteria in Milk,” International Journal of Food Micro-
biology, Vol. 87, No. 1-2, 2003, pp. 17-27.
doi:10.1016/S0168-1605(03)00044-8
[19] H. V. Atherton and J. A. Newlander, “Chemistry and
Testing of Dairy Products,” 4th Edition, AVI Publishing
Co. Inc., Madison, 1977.
[20] Y. G. Lee, H. Y. Wu, C. L. Hsu, C. J. Liang, H. D. Yuan.
“A Rapid and Selective Method for Monitoring the
Growth of Coliforms in Milk Using the Combination of
Amperometric Sensor and Reducing of Methylene Blue,”
Sensors and Actuators B: Chemical, Vol. 141, No. 2,
2009, pp. 575-580. doi:10.1016/j.snb.2009.06.028
[21] Q. Cai, K. Zeng, C. Ruan, T. A. Desai and C. A. Grimes,
“A Wireless, Remote Query Glucose Biosensor Based on
a pH-Sensitive Polymer,” Analytical Chemistry, Vol. 76,
No. 14, 2004, pp. 4038-4043. doi:10.1021/ac0498516
[22] L. G. Puckett, G. Barrett, D. Kouzoudis, C. A. Grimes
and L. G. Bachas, “Monitoring Blood Coagulation with
Magnetoelastic Sensors,” Biosensors and Bioelectronics,
Vol. 18, No. 5-6, 2003, pp. 675-681.
doi:10.1016/S0956-5663(03)00033-2
[23] C. Ruan, K. Zeng, O. K. Varghese and C. A. Grimes,
“Magnetoelastic Immunosensors: Amplified Mass Immu-
nosorbent Assay for Detection of Escherichia coli O157:
H7,” Analytical Chemistry, Vol. 75, No. 23, 2003, pp.
6494-6498. doi:10.1021/ac034562n
[24] R. Guntupalli, R.S. Lakshmanan, J. Hu, T. S. Huang, J. M.
Barbaree, V. Vodyanoy and B. A. Chin, “Rapid and Sen-
sitive Magnetoelastic Biosensors for the Detection of
Salmonella typhimurium in a Mixed Microbial Popula-
tion,” Journal of Microbiological Methods, Vol. 70, No. 1,
2007, pp. 112-118. doi:10.1016/j.mimet.2007.04.001
[25] S. Huang, S. Ge, L. He, Q. Cai and C. A. Grimes, “A
Remote-Query Sensor for Predictive Indication of Milk
Spoilage,” Biosensors and Bioelectronics, Vol. 23, No.
11, 2008, pp. 1745-1748. doi:10.1016/j.bios.2008.01.036
[26] P. D. Costa, N. J. De Andrade, F. J. V. Passos, S. C. C.
Brandão and C. G. F. Rodrigues, “ATP-Bioluminescence
as a Technique to Evaluate the Microbiological Quality of
Water in Food Industry,” Brazilian Archives of Biology
and Technology, Vol. 47, No. 3, 2004, pp. 399-405.
doi:10.1590/S1516-89132004000300010
[27] N. Ahmed, A. K. Mohanty, U. Mukhopadhyay, V. K.
Batish and S. Grover, “PCR-based Rapid Detection of
Mycobacterium Tuberculosis in Blood from Immuno-
competent Patients with Pulmonary Tuberculosis,” Jour-
nal of Clinical Microbiology, Vol. 36, No. 10, 1998, pp.
3094-3096.
[28] J. Haugen, R. Knut, S. Langsrud and S. Bredholt, “Ap-
plication of Gas-Sensor Array Technology for Detection
and Monitoring of Growth of Spoilage Bacteria in Milk:
A Model Study,” Analytica Chimica Acta, Vol. 565, No.
1, 2006, pp. 10-16. doi:10.1016/j.aca.2006.02.016
[29] M. Hébraud and P. Potier, “Cold Shock Response and
Low Temperature Adaptation in Psychrotrophic Bacte-
ria,” Journal of Molecular Microbiology and Biotech-
nology, Vol. 1, No. 2, 1999, pp. 211-219.
[30] Y. Liu, et al., “Prediction of Color, Texture, and Sensory
Characteristics of Beef Steaks by Visible and Near Infra-
red Reflectance Spectroscopy. A Feasibility Study,” Meat
Science, Vol. 65, No. 3, 2003, pp. 1107-1115.
doi:10.1016/S0309-1740(02)00328-5
[31] M. Lin, M. Mousavi, M. Al-Holy, A. G. Cavinato and B.
A. Rasco, “Rapid Near Infrared Spectroscopic Method for
the Detection of Spoilage in Rainbow Trout (Oncorhyn-
chus mykiss) Fillet,” Journal of Food Science, Vol. 71,
No. 1, 2006, pp. S18-S23.
doi:10.1111/j.1365-2621.2006.tb12400.x
[32] H. M. Al-Qadiri, M. Lin, M. A. Al-Holy, A. G. Cavinato
and B. A. Rasco, “Monitoring Quality Loss of Pasteur-
ized Skim Milk Using Visible and Short Wavelength
Near-Infrared Spectroscopy and Multivariate Analysis,”
Journal of Dairy Science, Vol. 91, No. 3, 2008, pp. 950-
958. doi:10.3168/jds.2007-0618
[33] N. Nicolaou, Y. Xu and R. Goodacre, “Fourier Transform
Infrared Spectroscopy and Multivariate Analysis for the
Detection and Quantification of Different Milk Species,”
Journal of Dairy Science, Vol. 93, No. 12, 2010, pp.
5651-5660. doi:10.3168/jds.2010-3619
[34] S. O. Yagoub, F. A. Bellow and I. El Zubeir, “Effect of
Temperature and Storage Period on the Constituents of
Milk Inoculated with Pseudomonas aeruginosa,” Re-
search Journal of Microbiology, Vol. 3, No. 1, 2008, pp.
30-34.
[35] E. Chiacchierini, F. D. Ascenzo, D. Restuccia and G.
Vinci, “Milk Soluble Whey Proteins: Fast and Precise
Determination with Dumas Method,” Analytical Letters,
Vol. 36, No. 11, 2003, p. 2473.
doi:10.1081/AL-120024336
[36] D. H. Kleyn, J. M. Lynch, D. M. Barbano, M. J. Bloom
and M. W. Mitchell, “Determination of Fat in Raw and
Processed Milks by the Gerber Method: Collaborative
Study,” Journal of AOAC International, Vol. 84, No. 5,
2001, pp. 1499-1508.
[37] E. H. Zacharov and M. Halpern, “Culturable Psychrotro-
phic Bacterial Communities in Raw Milk and Their Pro-
teolytic and Lipolytic Traits,” Applied and Environmental
Microbiology, Vol. 73, No. 22, 2007, pp. 7162-7168.
doi:10.1128/AEM.00866-07
[38] D. D. Muir, “The Shelf-Life of Dairy Products: 1. Factors
Influencing Raw Milk and Fresh Products,” International
Journal of Dairy Technology, Vol. 49, No. 1, 1996, pp.
24-32. doi:10.1111/j.1471-0307.1996.tb02616.x
[39] Tetra Pack International S.A., “Tetra Pak—About Proc-
essing, Packaging and Aseptic Technology,” 2013.
http://www.tetrapak.com/packages/aseptic_packages/page
s/default.aspx
[40] Evergreen Packaging, “Packaging Innovation,” 2013.
http://www.evergreenpackaging.com/fresh-by-design/inn
ovation

Milk Spoilage: Methods and Practices of Detecting Milk Quality 123
[41] C. Kouki, E. Sakin, Z. Jemai and Y. Dallery, “Assessing
the Impact of Perishability and the Use of Time Tem-
perature Technologies on Inventory Management,” In-
ternational Journal of Production Economics, Vol. 143,
No. 1, 2013, pp. 72-85.
[42] C. Fortin and H. L. Goodwin, “Valuation of Temp-
Time’s Fresh Check Indicator on Perishable Food Prod-
ucts in Belgium,” Selected Paper Prepared for Presenta-
tion at the Southern Agricultural Economics Association
Annual Meeting, Dallas, 2-5 February 2008, pp. 1-20.
[43] TempTime Corporation, “Products Overview,” 2009.
http://www.temptimecorp.com/publicpages/Products-Ove
rview.aspx