Electronic Nose (E-Nose) using

Micro Electro Mechanical Systems

(MEMS).

ABSTRACT

Many of us spend just as much time in cyber space touring the electronic landscapes of the Internet as

we spend offline. But for all of the time that we spend in front of our computer monitors, this virtual

world lacks many of the real world’s most precious attributes. One of the biggest drawbacks of the

cyber world is its lack of realism. Most of us are born with five senses, allowing us to see, hear,

touch, smell and taste. Yet, the Internet takes advantage of less than half of these.

When we log onto our computers, sight is probably the most obvious of the senses you see to collect

information. The Internet is almost completely vision-based.

While audio technologies like mp3 are also popular, but Internet is made up of words and pictures.

You can also throw in touch as the third sense in computer interaction, but that is mostly in terms of

interfacing by way of keyboard and mouse.

Since the beginning of Internet, software developers have chosen to ignore our senses of smell and

taste. However, there are at least some research works planning to awaken all our senses by bringing

digital odours to the Internet.

1. INTRODUCTION

Electronic/artificial noses are being developed as systems for the automated detection and

classification of odours, vapors, and gases. An electronic nose is generally composed of a chemical

sensing system (e.g., sensor array or spectrometer) and a pattern recognition system (e.g., artificial

neural network).

This Paper proposes developing electronic noses for the automated identification of volatile

chemicals for environmental and medical applications.

In this paper, I briefly describe an electronic nose, show some results from a prototype electronic

nose, and discuss applications of electronic noses in the defense, commercial, environmental, medical

and food industries.

Artificial odour detectors have been built since the 1950’s but without much success. In the past

few years, progression in chip technology and pattern recognition has made electronic noses available

for use in many different areas, from pollution detection to food processing.

The Institute of Olfactory Research at Warwick University developed the first prototype of

electronic nose in the mid-80s. An Electronic nose is a device used to analyze the content of air

through the classification of odours.

Although the electronic noses in use today are far from replacing the human olfactory system, the

possible uses for this technology are endless.

Electronic nose

Human noses are employed all over the world to test different products such as grains, wines,

cheeses, whiskey and fish are examined by human noses to determine their quality and freshness.

Perfumes and deodorants are also tested to see whether they are appealing to the nose.

The sense of smell is even used by doctors to help classify common disorders. Certain problems such as

pneumonia or diabetes give patient’s breath and bodily fluids characteristic odours that can be noticed

by a trained nose.

An electronic nose could be employed to do the same job with more possibilities.

THE BIOLOGICAL NOSE

To attempt to mimic the human apparatus, researchers have identified distinct steps that characterize

the way humans smell. It all begins with sniffing, which moves air samples that contain molecules of

odors past curved bony structures called turbinate. The turbinate create turbulent airflow patterns that

carry the mixture of volatile compounds to that thin mucus coating of the nose’s olfactory epithelium,

where ends if the nerve cells that sense odorants. The volatile organic compounds (VOCs) basic to

odors reach the olfactory epithelium in gaseous form or else as a coating on the particles that fill the

air we breathe. Particles reach the olfactory epithelium not only from the nostrils but also from the

mouth when food is chewed. As VOCs and particles carrying VOCs pass over the mucus membrane

lining the nose, they are trapped by the mucus and diffuse through to the next layer, namely, the

epithelium, where the sensory cells lie in wait. The cells are covered in multiple cilia- hair like

structures with receptors located on the cells outer membranes. Olfactory cells are specialized

neurons that are replicated approximately every 30 days.

The transformation of a molecule into an odor begins when this odorant molecule, as it is called,

binds to a receptor protein. The event initiates a cascade of enzymatic reactions that result in

depolarization of the cell’s membrane. (Ion pumps within the cell’s membrane keep the cell polarized

in its rest, or steady state, with a typical rest potential of about 90 mV across the membrane). There

are more than 100 million protein receptors in all and perhaps 1000 types. For example, one receptor

type is sensitive to a small subset of odorants, one of which is the organic compound octanal. The

sensory cells in the epithelium respond by transmitting signals along neural “wires” called axons.

Such an axon first traverses a small hole in a bony structure in the base of the skull, known as the

cribriform plate. Then the rest of the neuron wends its way to the brain’s olfactory bulb, where it

terminates in a cluster of neural networks called glomeruli. The 2000 or so glomeruli of the olfactory

bulb represent the first tier of central odor information processing. All sensory neurons containing a

specific odor an receptor are thought to converge on two or three

glomeruli in the olfactory bulb. Note that olfactory sensory neurons in the epithelium can each

respond to nose than one odorant. It is therefore the pattern of response acrossmultiple glomeruli that

codes olfactory quality.

SIGNAL PROCESSING AND PATTERNRECOGNITION

The task of an electronic nose is to identify an odorant sample and perhaps to estimate its

concentration. The means are signal processing and pattern recognition. For an electronic nose

system this two steps may be subdivided into four sequential stages. They are preprocessing, feature

extraction, classification and decision making. But first a data base of the expected odorant must be

compiled, and sample must be presented to the nose’s sensor array. Preprocessing compensates for

sensor drift compress the transient response of the sensor array, and reduces sample to sample

variations. Typical techniques are

Mani pulation of sensor base lines, normalization of sensor response ranges of all the sensors in an

array (the normalization constant may some times be used to estimate the odorant concentration), and

compression sensor transients. Feature extraction has two purposes; they are to reduce the

dimensionality of the measurement space, and to extract information relevant for pattern recognition.

To illustrate, in an electronic nose with 32 sensors, the measurement space has 32 dimensions. This

space can cause statistical problem if odor database contains only a few examples, typical in pattern

recognition applications because of the cost of data collection. Further more, since the sensors have

overlapping sensitivities there is high degree of redundancy in these 32 dimensions. Accordingly is it

convenient to project the 32 on to a few informative and independent axes. This low dimensional

projection (typically 2 or 3 axes) has the added advantage that it can be more readily inspected

visually.

Feature extraction is generally performed with linear transformations such as the classical principal

component analyses (PCA) and linear discriminate analysis (LDA). PCA finds projections of

maximum variance and is the most widely used linear feature extraction techniques. But it is not

optimal for classification since it ignores the identity (class label) of the odor examples in the

database. LDA, on the other hand, looks at the class label of each example. Its goal is to find

projections that maximize the distance between examples from different odorants yet minimize the

distance between examples of the same odorant. As in example, PCA may do better with a projection

that contain high variance random noise whereas LDA may do better with a projection that contain

subtle, but maybe crucial, odor discriminatory information. LDA is therefore more appropriate for

classification purposes. Several research groups have recently adopted some nonlinear transforms,

such as Sammon nonlinear maps and Kohonen self organizing maps. Sammon maps attempt to find a

2D or 3D mapping the preserves the distance between pairs of examples on the original 32

dimensional space. Kohonen maps project the 32 dimensional space onto a two dimensional mesh of

processing elements called neurons.

Neighboring neurons are trained to respond to similar types of stimuli (odorants), a self organizing

behavior motivated by neuro biological considerations. Neither of these techniques makes use of

class labels, so they are not optimal for pattern classification. Once the odor examples have been

projected on an appropriate low dimensional space the classification stage can be trained to identify

the patterns that are representative of each odor, when presented with an unidentified odor, the

classification stage will be able to assign to it a class label (identify the odorant) by comparing its

pattern with those complied during training.

The classical methods of performing the classification task are K nearest neighbor (KNN) Bayesian

classifiers, and Artificial Neural Network(ANN). An Artificial Neural Network (ANN) is an

information processing paradigm that is inspired by the way nervous systems, such as the brain,

process information.

They key elements of this paradigm is the novel structure of the information processing system. It is

composed of a large number of highly interconnected processing elements (neurons) working in

unison to solve specific problems. ANNs, like people, learn by eample. An ANN is configured for a

specific application, such as pattern recognition or data classification through a learning process.

Learning in biological systems involves adjustments to the synaptic connections that exist between

the neurons.

This is true of ANNs as well. For the electronic nose, the ANN learns to identify the various chemical

or odors by example. A typical ANN classifier consists of two or more layers of neurons that are

connected with synaptic weights-real number multiplier that connect that output of neuron to the

inputs of neurons in the next layer.

During the training the Ann adapts the synaptic weights to learn the pattern of different odornants.

After training when presented with an unidentified odorant the ANN feeds its pattern through the

different layers of neurons and assigns the class label that provides the largest response. ANNs have

been applied to and increasing number of real world problems of considerable complexity.

Their most important advantage is in solving problems that are too complex for conventional

technologies; that is problems that do not have an algorithmic solution or for which an algorithmic

solution is too complex to be found.

In general, because of their abstraction from the biological brain, ANNs are well suited to problems

that people are good at solving but computers are not. These problems include pattern recognition

and forecasting. However, unlike the human capability in pattern recognition, The ANNs capability is

not affected by factors such as fatigue, working conditions, emotional state and compensation.

ELECTRONIC NOSE PRINCIPLES

Enter the gas sensors of the electronic nose. This speedy, reliable new technology undertakes what till

now has been impossible – continuous real monitoring of odor at specific sites in the field over hours,

days, weeks or even months.

An electronic device can also circumvent many other problems associated with the use of human

panels. Individual variability, adaptation (becoming less sensitive during prolonged exposure),

fatigue, infections, mental state, subjectivity, and exposure to hazardous compounds all come to

mind. In effect, the electronic nose can create odor exposure profiles beyond the capabilities of the

human panel or GC/MS measurement techniques.

The electronic nose is a system consisting of three functional components that operate serially on an

odorant sample- a sample handler, an array of gas sensors, and a signal processing system. The output

of the electronic nose can be the identity of the odorant, an estimate of the concentration of the

odorant, or the characteristic properties of the odor as might be perceived by a human. Fundamental

to the artificial nose is the idea that each sensor in the array has different sensitivity.

For example, odorant No. 1 may produce a high response in one sensor and lower responses in

others, whereas odorant No. 2 might produce high readings for sensors other than the one that “took”

to odorant No.1.

What is important is that the pattern of response across the sensors is distinct for different odorants.

This distinguishability allows the system to identify an unknown odor from the pattern of sensor

responses. Each sensor in the array has a unique response profile to the spectrum of odorants under

test. The pattern of response across all sensors in the array is used to identify and/or characterize the

odor.

Sensing an odorant

In a typical electronic nose, an air sample is pulled by a vacuum pump through a tube into a small

chamber housing the electronic sensor array. The tube may be made of plastic or a stainless steel.

Next, the sample– handling unit exposes the sensors to the odorant, producing a transient response as

the VOCs interact with the surface and bulk of the sensor’s active material. (Earlier, each sensors has

been driven to a known state by having clean, dry air or some other reference gas passed over its

active elements.) A steady state condition is reached in a few seconds to a few minutes, depending on

the sensor type. During this interval, the sensor’s response is recorded and delivered to the signal

processing unit. Then, a washing gas such as an alcohol vapor is applied to the array for a few

seconds to a minute, so as to remove the odorant mixture from the surface and bulk of the sensor’s

active material. (Some designers choose to skip this washing step) Finally, the reference gas is

applied to the array, to prepare it for a new measurement cycle. The period during which the odorant

is applied is called the response time of the sensor array. The period during which the washing and

reference gases are applied is termed the recovery time.

HISTORY

Scientist Alexander Graham Bell popularized the notion that it was difficult to measure a smell, and

in 1914 said the following:

Did you ever measure a smell? Can you tell whether one smell is just twice strong as another? Can

you measure the difference between one kind of smell and another? It is very obvious that we have

very many different kinds of smells, all the way from the odor of violets and roses up to asafetida.

But until you can measure their likeness and differences, you can have no science of odor. If you are

ambitious to find a new science, measure a smell.

—Alexander Graham Bell, 1914

In the decades since Bell made this observation, no such science of odor materialized, and it was not

until the 1950s and beyond that any real progress was made.

WORKING PRINCIPLE

The electronic nose was developed in order to mimic human olfaction that functions as a non-

separative mechanism: i.e. an odor / flavor is perceived as a global fingerprint. Essentially the

instrument consists of head space sampling, sensor array, and pattern recognition modules, to

generate signal pattern that are used for characterizing odors.

Electronic noses include three major parts: a sample delivery system, a detection system, a computing

system.

The sample delivery system enables the generation of the headspace (volatile compounds) of a

sample, which is the fraction analyzed. The system then injects this headspace into the detection

system of the electronic nose. The sample delivery system is essential to guarantee constant operating

conditions.

The detection system, which consists of a sensor set, is the "reactive" part of the instrument. When in

contact with volatile compounds, the sensors react, which means they experience a change of

electrical properties.

In most electronic noses, each sensor is sensitive to all volatile molecules but each in their specific

way. However, in bio-electronic noses, receptor proteins which respond to specific odor molecules

are used. Most electronic noses use sensor arrays that react to volatile compounds on contact: the

adsorption of volatile compounds on the sensor surface causes a physical change of the sensor. A

specific response is recorded by the electronic interface transforming the signal into a digital value.

Recorded data are then computed based on statistical models.

Bio-electronic noses use olfactory receptors - proteins cloned from biological organisms, e.g.

humans, that bind to specific odor molecules. One group has developed a bio-electronic nose that

mimics the signaling systems used by the human nose to perceive odors at a very high sensitivity:

femtomolar concentrations.[6]

The more commonly used sensors for electronic noses include

Metal–oxide–semiconductor (MOSFET) devices - a transistor used for amplifying or

switching electronic signals. This works on the principle that molecules entering the sensor

area will be charged either positively or negatively, which should have a direct effect on the

electric field inside the MOSFET. Thus, introducing each additional charged particle will

directly affect the transistor in a unique way, producing a change in the MOSFET signal that

can then be interpreted by pattern recognition computer systems. So essentially each

detectable molecule will have its own unique signal for a computer system to interpret.

conducting polymers - organic polymers that conduct electricity.

quartz crystal microbalance - a way of measuring mass per unit area by measuring the change

in frequency of a quartz crystal resonator. This can be stored in a database and used for future

reference.

surface acoustic wave (SAW) - a class of microelectromechanical systems (MEMS) which

rely on the modulation of surface acoustic waves to sense a physical phenomenon.

Some devices combine multiple sensor types in a single device, for example polymer coated QCMs.

The independent information leads to vastly more sensitive and efficient devices.

In recent years, other types of electronic noses have been developed that utilize mass spectrometry or

ultra-fast gas chromatography as a detection system.

The computing system works to combine the responses of all of the sensors, which represents the

input for the data treatment.

This part of the instrument performs global fingerprint analysis and provides results and

representations that can be easily interpreted. Moreover, the electronic nose results can be correlated

to those obtained from other techniques (sensory panel, GC, GC/MS). Many of the data interpretation

systems are used for the analysis of results. These systems include artificial neural network (ANN),

fuzzy logic, pattern recognition modules, etc.

How to perform an analysis

As a first step, an electronic nose needs to be trained with qualified samples so as to build a database

of reference. Then the instrument can recognize new samples by comparing volatile compounds

fingerprint to those contained in its database. Thus they can perform qualitative or quantitative

analysis. This however may also provide a problem as many odours are made up off multiple

different molecules, this may be possibly wrongly interpreted by the device as it will register them as

different compounds, resulting in incorrect or inaccurate results depending on the primary function of

a nose.

2. How an e-nose works?

The two main components of an electronic nose are the sensing system and the automated pattern

recognition system. The sensing system can be an array of several different sensing elements (e.g.,

chemical sensors), where each element measures a different property of the sensed chemical, or it can

be a single sensing device (e.g., spectrometer) that produces an array of measurements for each

chemical, or it can be a combination.

Each chemical vapor presented to the sensor array produces a signature or pattern characteristic of

the vapor. By presenting many different chemicals to the sensor array, a database of signatures is

built up. This database of labeled signatures is used to train the pattern recognition system. The goal

of this training process is to configure the recognition system to produce unique classifications of

each chemical so that an automated identification can be implemented.

The e-noses work on the same principle as the human nose. Humans detect odours using up to 650

types of receptors found on cells high in the nasal passages. Scientists are still not sure how the

human nose processes smells but its known that the receptors generate ‘smell prints’ of odours they

receive, which then passes on this information to the brain.

Multivariate analysis

Multivariate data analysis generally involves data

reduction, it reduces high dimensionality in a multivariate problem where variables are partly

correlated (e.g. sensors with overlapping sensitivities), allowing the information to be displayed in a

smaller dimension (typically two or three) (45, 46). There are many multivariate techniques to choose

from. A subset called pattern recognition (PARC) techniques are routinely used in electronic nose

data analysis.

Supervised/unsupervised

The PARC methods can be divided into unsupervised or untrained techniques, and supervised or

trained techniques. Unsupervised learning methods are generally used in exploratory data analysis

because they attempt to identify a gas mixture without prior information on the nature of the samples.

These techniques, which include PCA, CA and multi-dimensional scaling, are useful when no

example of different sample groups is available, or when hidden relationships between samples or

variables are suspected (3, 45). Conversely, supervised learning techniques classify an odour by

developing a mathematical model relating training data, i.e. samples with known properties, to a set

of given descriptors. Test samples are then evaluated against a knowledge base and predicted class

membership is deduced. These methods enable the system to reduce parameters other than volatile,

for example temperature and humidity, and train a system to look only at particular combinations of

sensors to measure a given odour. Generally, electronic nose data are best processed using trained

techniques such as CDA, FW, ANN or RBF (3).

Linear/non-linearThe above multivariate analyses are all linear PARC methods where a model is calculated using

linear combinations of input data (30). Most sensors have a non-linear response versus concentration;

however, these techniques work well if a low concentration of volatiles ensures an approximately

linear response. In addition, the use of pre-processing algorithms, such as averaging, linearization or

normalization, improve the performance of these analytical tec

With e-noses, the process is same but sensors take the place of human cell receptors and a

microprocessor takes the place of the brain. The e-nose also creates smell prints of the odours it

receives, which are then stored in a database. If it receives the same imprint in the future, it will

recognize that odour from its database and inform the user of the smell.

The quantity and complexity of the data collected by sensors array can make conventional

chemical analysis of data in an automated fashion difficult. One approach to chemical vapor

identification is to build an array of sensors, where each sensor in the array is designed to respond to

a specific chemical. With this approach, the number of unique sensors must be at least as great as the

number of chemicals being monitored. It is both expensive and difficult to build highly selective

chemical sensors.

Artificial neural networks (ANN’s), which have been used to analyze complex data and to

recognize patterns, are showing promising results in chemical vapor recognition. When an ANN is

combined with a sensor array, the number of detectable chemicals is generally greater than the

number of sensors. Also, less selective sensors which are generally less expensive can be used with

this approach.

The sensors basically measure the change in voltage due to presence of certain chemicals. The

chemicals in the air change the oxygen content over the sensors. Change in the oxygen content

changes the resistance across the sensor, which can be measured as a voltage drop from the normal or

standardized conditions. This analog signal must be translated into a digital signal by an analog-to-

digital converter (ADC) for the computer to understand the information. The number of odour

signatures the system can recognize depends on the number of sensors used and the grey levels in the

converter.

Electronic noses that incorporate ANN’s have been demonstrated in various applications. Some of

these applications will be discussed later in the paper. Many ANN configurations and training

algorithms have been used to build electronic noses.

2.1. Sensors built with Micro Electro Mechanical Systems.

In less than 20 years, MEMS (micro electro-mechanical systems) technology has gone from an

interesting academic exercise to an integral part of many common products. But as with most new

technologies, the practical implementation of MEMS technology has taken a while to happen.

The sensors are conductive polymer – carbon black composite films, which swell reversibly and

induce a resistance change upon exposure to a wide variety of gases. The ability to monitor and

detect various gases is important for many applications.

Monitoring and determining the constituents of a sample gas typically involves collecting samples

and analyzing them in a gas chromatograph-mass spectrometer (GC-MS) of significant size and cost

Although GC-MS systems work very well, many applications need sensor systems that are smaller,

more portable, cheaper, and even disposable. Conductive polymer – carbon black composite films

have been used as a sensing medium in several gas sensors or “electronic noses” .These composite

films swell reversibly when exposed to various gases. This swelling induces a resistance change in

the composite film, which can be measured when the film is deposited across two metal leads.

Compared to conventional chemical sensors, which uses a specific receptor that selectively

responds to a single analyte of interest , the polymer composite film is not specific to any one

particular gas. However, when it is used in an array, with each sensor containing a different polymer

composite film, gases and gas mixtures can be identified by the pattern of response of the array. This

allows a much more general-purpose chemical gas sensor capable of broadly detecting and

identifying various constituents.

These polymers – carbon black composite films have been deposited on co-fired ceramic substrates

, glass slides , and on silicon . Most of these sensors have large-area composite film deposits (>mm 2).

During deposition, the composite film is dissolved into a solvent mixture, which evaporates and

leaves behind the thin sensing film.

Using micromachining technology, we have been able to develop smaller polymer-based chemical

gas sensor arrays on silicon. Specifically, we have designed micromachined reservoirs or “wells” to

contain the large liquid volume present during deposition.

The “well” allows the polymer – carbon black film to be placed reproducibly in a specific and

well-constrained area. The 3-dimensional structure of the “well” can also afford a larger exposure

area for sensing while still minimizing the chip area used by each element of the array. These “wells”

can be post processed onto silicon CMOS chips, which would allow for integration of on-chip

electronics for measurement, signal processing, and analysis.

The “well” sizes that we have fabricated and tested are 500×500 µm, 250×250 µm, and 100×100

µm.

3. What is an Artificial Neural Network (ANN)?

An artificial neural network is an information-processing paradigm inspired by the way biological

nervous systems, such as the brain, process information.

It has a large number of highly interconnected processing elements (neurons) working in unison to

solve specific problems.

Like Humans, ANN’s learn by example. An ANN is configured for an application such as

identifying chemicals vapors through a learning process. Learning in biological systems involves

adjustments to the synaptic connections.

This is true of ANN’s as well. For the electronic nose, the ANN learns to identify the various

chemicals or odours by example.

ANN’s have been applied to an increasing number of real-world problems of considerable

complexity. These can solve problems that are too complex for conventional technologies i.e.

problems that do not have an algorithmic solution or for which an algorithmic solution is too

complex to be found.

In general, because of their abstraction from the biological brain, ANN’s are well suited to problems

that people are good at solving but computers are not; for example, pattern recognition and

forecasting.

However, unlike the humans, the pattern recognition capability of ANN’s is not affected by fatigue,

working conditions, emotional state and compensation.

4. Prototype Electronic Nose

A Prototype electronic nose, shown 1inFigure, is composed of an array of tin-oxide vapor sensors,

a humidity sensor, and a temperature sensor coupled with an ANN. During operation a chemical

vapor is blown across the array, the sensor signals are digitized and fed into the computer, and the

ANN (implemented in software) then identifies the chemical. This identification time is limited only

by the response time of the chemical sensors, which is in order of seconds. This prototype nose has

been to identify common household chemicals by their odour.

The tin-oxide sensors are commercially available sensors. Although each sensor is designed for a

specific chemical, each responds to a wide variety of chemicals. Collectively, theses sensors respond

with unique signatures (patterns) to different chemicals. During the training process, various

chemicals with known mixture are presented to the system. By training on samples of various

chemicals, the ANN learns to recognize the different chemicals.

Gas sensors tend to have very broad selectivity, responding to many different substances. This is a

disadvantage in most applications, but in electronic nose, but is a definite advantage. Although every

sensor in an array may respond to a given chemical, theses responses will usually be different.

This prototype nose has been tested on a variety of household and office supply chemicals including

acetone, ammonia, ethanol, glass cleaner, contact cement, correction fluid, iso-propanol, lighter fluid,

methanol, rubber cement and vinegar. For results shown in paper, three of these chemicals were used:

acetone, benzene and chloroform. The figure illustrates the responses of the sensors numbered from 1

to 9 for a variety of test chemical presented to the ANN’s. The ANN was able to correctly classify the

test samples with only small residual errors.

From the responses of the sensors to the analysis, one can easily see that the individual sensors in the

array are not selective. In addition, when a mixture of two or more chemicals is presented to the

sensor array, the resultant pattern (sensor values) may be even harder to analyze. Thus, analyzing the

sensor responses separately may not be adequate to yield the classification accuracy achieved by

analyzing the data in parallel.

Why do we need to use an electronic nose?

We use a very sensitive and powerful instrument to detect odours - our noses!

We also have a very powerful computer to interpret all of the data we receive - our brains!

The problem with using our nose and brain, as an analytical tool, is that they provide a subjective

analysis. For example, country dwellers tend to find agricultural smells as offensive as those who live

in towns. The electronic nose mimics some of the characteristics of the human nose but is not

subjective. The model we are using is a new portable electronic nose (Cyranose 320) supplied by our

research partner Cyrano Sciences. It contains 32 individual sensors, each of which is made from a

different organic polymer-carbon black combination. When the sensors are exposed to an odour they

expand slightly and this causes a change in their electrical conductivity. This change, which is usually

different for each sensor, is recorded. The pattern of response from the 32 sensors varies for different

odours. Sophisticated computer programs are used to compare the pattern from an unknown material

with a library of known patterns. An identification of a new sample can be made if it matches any of

the known patterns in the library.

Our aim is to test this new sensor technology in a practical application to help solve problems

associated with odours released from industrial and other sites and which cause a nuisance to the

general public.

Progress Report

We have investigated many issues associated with the use of electronic noses for environmental

sampling, including selectivity, sensitivity, the effects of humidity and of sensor drift. We have also

carried out a study on agricultural wastes. We were able to use the e-nose to distinguish between five

different agricultural wastes over a five month period.

The human nose is incredibly sensitive to odorous compounds - often detecting concentrations of a

few parts per billion. However, the Cyranose 320 is not as sensitive as the human nose, and, in order

to detect odours at such low concentrations, an odorous sample needs to be preconcentrated before it

is sampled by the nose. The preconcentration system used is the EDU by Airsense Analytics. This

system was chosen as it is small, light weight and can be powered by a battery.

5. Applications

5.1. Electronic Noses for Environmental Monitoring.

Enormous amounts of hazardous waste (nuclear, chemical, and mixed wastes) were generated by

more than 40 years of weapons production in the U.S. Department of Energy’s Weapons Complex.

The Pacific Northwest National Laboratory is exploring the technologies required to perform

environmental restoration and waste management in a cost effective manner. This effort includes the

development of portable, inexpensive systems capable of real-time identification of contaminants in

the field. Electronic noses fit this category (As part of this effort, ANN’s are being combined with

chemical sensor arrays and spectrometers for use in prototype electronic noses).

Environmental applications of Electronic noses include analysis of fuel mixtures, detection of oil

leaks, testing ground water for odours, and identification of household odours. Potential applications

include identification of toxic wastes, air quality monitoring and monitoring factory emissions.

5.2. Electronic Noses for Medicine

Because the sense of smell is an important sense to the physician, an electronic nose has

applicability as a diagnostic tool. An electronic nose can examine odours from the body (e.g., breath,

wounds, body fluids, etc) and identify possible problems. Odours in the breath can be indicative of

gastrointestinal problems, sinus problems, infections, diabetes, and liver problems. Infected wounds

and tissues emit distinctive odours that can be detected by an electronic nose for examining wound

infections is being tested at South Manchester University Hospital.

A More futuristic application of electronic noses has been recently proposed for tele-surgery.

While the inclusion of visual, aural, and tactile senses into tele-present systems is widespread, the

sense of smell has been largely ignored. An electronic nose will potentially be a key component in an

olfactory input to tele-present virtual reality systems including tele-surgery.

The electronic nose would identify odours in the remote surgical environment. These identified

odours would be electronically transmitted to another site where an odour generation system would

recreate them.

5.3. Electronic Noses for the Food Industry.

Currently, the biggest market for electronic noses is the food industry. Applications of electronic

noses in the food industry include quality assessment in food production, inspection of food quality

by odour, control of food cooking processes, inspection of fish, monitoring the fermentation process,

checking rancidity of mayonnaise, verifying if orange juice is natural, monitoring food and beverage

odours, grading whisky, inspection of beverage containers, checking plastic wrap for containment of

onion odour, and automated flavor control to name a few.

In some instances electronic noses can be used to augment or replace panels of human experts. In

other cases, electronic noses can be used to reduce the amount of analytical chemistry that is

performed in food production especially when qualitative results will do.

A number of potential applications of the electronic nose in food industry have been reported, such

as for quality estimation of ground meat (Winquist et al., 1993), detection of boar taint in meat

(Bourrounet et al., 1995), detection of gender differences in meat products (Berdague and Talou,

1993), estimation of fish freshness (Scweizer-Berberich et al., 1994. Winquist et al., 1995),

evaluation of shelf life of fresh vegetables (Riva et al., 2001), to follow beer production (Pearce et

al., 2003) and to predict the degree of mouldy/musty odour in cereal (Borjesson et al., 1996).

5.5. Electronic Noses for Business and other applications.

In the future, bomb squads won’t need dogs. They’ll have electronic sniffers to find explosives.

The portable noses will also uncover sub-standard food, and sound an alarm if a car’s brake fluid

smells deficient. Working with Caltech’s center for Neuromorphic Systems Engineering, an NSF

Engineering Research Center (ERC), Researcher’s say that this new approach to artificial olfactory

sensing. Most other chemical sensors are keyed to specific chemicals.

When those chemicals are absent, the noses don’t work. The Caltech nose however will learn, like

a dog, it will memorize scents and react to something it hasn’t smelled before. This versatility

intrigues NASA. No one knows what new gas may appear inside the pace shuttle, but the Caltech

nose could learn what the spacecraft smells like under normal conditions, and then ring a bell if the

odours change.

Perfume makers are already using nose-machines to protect their patented smells against fake-

fragrance merchants. Inspectors are using high-tech snout to resolve disputes with fishermen over

grading of their catch.

More exciting are the possible medical applications: Warwick University scientists are researching

the use of electronic noses to diagnose illness by smelling patient’s breath, and have recently been

awarded a grant from the European Union (EU) to investigate the possibility of installing tiny

electronic nose in phone receivers, so patients can simply breathe into the phone and wait for

diagnosis. A similar smell-transmission device may soon allow surfer’s on the internet to ‘wake up

and smell the coffee’ quite literally.

Researchers are investigating the use of breath analysis to identify the stages of female menstrual

cycle: the ability of electronic noses to detect ovulation could benefit both fertility and birth control.

Our unique body odour may become an alternative form of identification, signaling the end of credit

card fraud, forgotten or misappropriated personal identification number (PIN), fake ID cards, etc.

The association for payment clearing services, an organization set up to find solutions to these

problems, is investigating the use of electronic nose in banks, and companies may soon be able to

replace security entry systems involving cards and codes with a device that recognizes each

employee’s personal odour.

5.4. E-mail the Smells.

In Oakland, California, DigiScents Inc. is developing a device called ‘iSmell’ that transmits

digitized smells through your computer. The prototype of the iSmell personal scent synthesizer will

connect to your PC through a Serial or Universal Serial Bus (USB) port. It can be plugged into any

ordinary electrical outlet.

Here’s how it works: DigiScents prepares an index of thousands of smells based on their chemical

structure and their place on the smell spectrum. Each scent is then coded and digitized into a small

file. The digital file is embedded in web content or e-mail. When a user requests or triggers the file

by clicking or opening the e-mail, a small amount of the aroma is emitted by the device in the direct

vicinity of the user.

The iSmell can create thousands of everyday scents with a small cartridge that contains 128

primary odours. These primary odours are mixed together to generate other smells that closely

replicate common natural and manmade odours. The scent cartridge, like a printer’s toner cartridge,

will have to be replaced periodically to maintain the scent accuracy.

DigiScents has partnered with several Web, Interactive Media and gaming companies to bring

scents to your computer. Real Networks plans to make DigiScents Scentstream software available to

its more than 115 million RealPlayer users.

5.5. Print Flavors

Tri Senx is planning to take you one step further by allowing you to not only download scents but

also print flavors that can be tasted. It has developed a technology that allows users to print smells

onto thick fiber-paper sheets and taste specific flavors by licking the paper coated with the smell.

The SENX is a printer like desktop device that will produce smells based on data programmed into

Webpage. The fragrances and aromas are stored in a disposable cartridge within the SENX. This

cartridge has 20 chambers, each holding a distinct scent. Thousands of smells can be created with a

20 chamber cartridge. The SENX measures 14x20x6.4 cm and is powered by a 6V DC rechargeable

battery. It plugs into the open external COM port on your computer.

5.6. Using E-Smells.

You will be able to do more than attach e-smells. Imagine watching The Patriot on your DVD

player with DigiScents device plugged into it: as the Colonial army’s cannons blast, you can smell

the gunpowder. The whole idea here is to increase the realism and enhance the viewing of your

favorite movies. The same type of effect could be created for your favorite video games.

While consoles like PlayStation3 are designed to enhance realism of video game graphics, a digital

scent synthesizer could take games to a whole new level. Imagine smelling the bad guy who is

approaching before you actually see him. Developers of racing games could embed the smell of burnt

rubber or gasoline to make their games more realistic.

Before being attached to movies and games, Internet odours are likely to permeate through Internet

Advertising. Advertisers can use novelty of digital scents to peddle their products now. Coca-Cola

could embed their cola smell into banner ads, which could be triggered by a user scrolling over the

ad.

Consumers may also benefit from this aromatic technology. With online spending on the rise,

shoppers will be able to sample some goods that they buy including flowers, candy, coffee and other

food products.

Electronic nose instruments are used by research and development laboratories, quality control laboratories and process & production departments for various purposes:

In quality control laboratories for at line quality control such as

Conformity of raw materials, intermediate and final products

Batch to batch consistency

Detection of contamination, spoilage, adulteration

Origin or vendor selection

Monitoring of storage conditions.

In process and production departments

Managing raw material variability

Comparison with a reference product

Measurement and comparison of the effects of manufacturing process on products

Following-up cleaning in place process efficiency

Scale-up monitoring

Cleaning in place monitoring.

Possible and future applications in the fields of health and security

The detection of dangerous and harmful bacteria, such as software that has been specifically developed to recognise the smell of the MRSA (Methicillin-resistant Staphylococcus Aureus).[12] It is also able to recognise methicillinsusceptible S. aureus (MSSA) among many other substances. It has been theorised that if carefully placed in hospital ventilation systems, it could detect and therefore prevent contamination of other patients or equipment by many highly contagious pathogens.

The detection of lung cancer or other medical conditions by detecting the VOC's (volatile organic compounds) that indicate the medical condition.[13][14]

The quality control of food products as it could be conveniently placed in food packaging to clearly indicate when food has started to rot or used in the field to detect bacterial or insect contamination.[15]

Nasal implants could warn of the presence of natural gas, for those who had anosmia or a weak sense of smell.

Possible and future applications in the field of crime prevention and security

The ability of the electronic nose to detect odourless chemicals makes it ideal for use in the police force, such as the ability to detect drug odours despite other airborne odours capable of confusing police dogs. However this is unlikely in the mean time as the cost of the electronic nose is too great and until its price drops significantly it is unlikely to happen.

It may also be used as a bomb detection method in airports. Through careful placement of several or more electronic noses and effective computer systems you could triangulate the location of bombs to within a few metres of their location in less than a few seconds.

In environmental monitoring

For identification of volatile organic compounds in air, water and soil samples.[16]

For environmental protection.[17]

Various application notes describe analysis in areas such as flavor and fragrance, food and beverage, packaging, pharmaceutical, cosmetic and perfumes, and chemical companies. More recently they can also address public concerns in terms of olfactive nuisance monitoring with networks of on-field

devices.[18][19] Since emission rates on a site can be extremely variable for some sources, the electronic nose can provide a tool to track fluctuations and trends and assess the situation in real time. It improves understanding of critical sources, leading to pro-active odor management. Real time modeling will present the current situation, allowing the operator to understand which periods and conditions are putting the facility at risk. Also, existing commercial systems can be programmed to have active alerts based on set points (odor concentration modeled at receptors/alert points or odor concentration at a nose/source) to initiate appropriate actions.

6. Key Benefits.

The Electronic noses available so far are not as sensitive as human nose, but these offer significant

advantages: These do not get bored with repetitive smelling tasks, or de-sensitized through

habituation to particular odours. Unpleasant smells such as industrial chemicals and sewage do not

make electronic sniffers feel sick, and their performance on smelling tasks does not fluctuate

according to mood, hormonal cycles or other unpredictable human factors. For most tasks, one of the

main advantages of electronic noses is their lack of emotional response to odours, although in future

high-tech noses may be developed that mimic human emotions (perhaps for perfume makers to test

the effects their products).

One of the many problems with using the human olfactory system as a tool is that it is extremely

subjective. Different people are affected in different ways by similar odours. An electronic nose

could solve this problem by setting standard for certain smells. Another problem that electronic nose

could solve is the health risk associated with smelling certain chemicals.

The toxicity of certain chemicals could prove harmful to those who try determining an odour using

their nose. In addition electronic noses could be taken into areas of extreme temperatures, inside the

body, oil rigs, gasoline tanks or sewer systems, or even another planet.

However, e-noses don’t come without disadvantage. First, these are temperature-sensitive an

increase or decrease in the temperature lessens their effect. The atmospheric moisture and humidity

also result in improper working of the e-nose.

7. Present status and Future.

E-nose is already available in the market. The two main manufacturers are EST and Cyrano

Sciences, both USA based companies. Taking e0noses a step further, illumine, another USA based

company is working on optical nose. Meanwhile Caltech’s ERC are working on the computer design

of the nose. The completed sniffer will be made up of 10,000 sensors and fit on a 1cm2 chip.

Eventually, many questions concerning the human olfactory system will be answered and perhaps

artificial solutions will be found for people with olfactory problems.

6. ELECTRONIC NOSE INSTRUMENTATION

The basic element of a generalized electronic instrument system to measure

odours is shown schematically in the figure. First there is an odour from the source

material to the sensor chamber.

There are tow main ways in which the odour can be delivered to the sensor

chamber, namely head space sampling and flow injection. In head space sampling, the

head space of an odorant material is physically removed from a sample vessel and

inserted into the sensor chamber using either a manual or automated procedure.

Alternatively, a carrier gas can be used to carry the odorant from the sample vessel into

the sensor by a method called flow injection. The sensor chamber houses the array of

chosen odour sensors, e.g. Semi conducting polymer chemo resisters, etc. The sensor

electronic not only convert the chemical signal into an electrical signal into an

electrical signal into an electrical signal into an electrical signal but also, usually,

amplify and condition it. This can be done using conventional analogue electronic

circuitry (e.g. operational amplifiers) and the output is then a set of an analogue

outputs, such as 0 to 5v d.c. although a 4 to 20mA d.c. current output of preferable if

using a long cable. The signal must be converted into a digital converter (e.g. a 12 – bit

converter) followed by a multiplexer to produce a digital signal which either interfaces

to a serial port on the microprocessor (e.g. RS - 232) or digital bus (e.g. GPIB). The

microprocessor (e.g. an Intel 486 or Motorola 68HC11) is programmed to carry out a

number of tasks.

CONCLUSION

An ‘electronic nose’ is a system originally created to mimic the function of an animal nose. However,

this analytical instrument is more a ‘multi-sensor array technology’ than a real ‘nose’. Whatever the

sensor technology, it is still far from the sensitivity and selectivity of a mammalian nose. Therefore,

its aim is not to totally replace either the human nose or other analytical methods. A sensory panel is

necessary to define the desired product quality which can then be used to train the system. Traditional

analytical methods qualitatively or/and quantitatively why one food sample differs from others. The

‘electronic nose’ can only perform quick ‘yes or no’ tests in comparison to other products. It could

occasionally replace sensory analysis and even perform better than a sensory panel in routine work,

or in cases where non-odorous or irritant gases need to be detected. Therefore, an ‘electronic nose’

can be regarded as an interesting tool for a quick quality test in various food applications. However,

before it can be treated as a completely reliable, industrial instrument, much improvement is still

needed, such as a reduction of the dead volume, development of calibration methods, etc. As current

knowledge does not allow the replacement of the human nose, constructors tend to compensate by

integrating several sensor technologies into one instrument. However, one single instrument to be

used in every possible application would be over-complicated due to the large number of sensors and

time consuming statistical analysis. The trend is to create a system for one specific application. This

means that a compact and portable instrument would be desirable (108).

8. References.

[1] Berdague J.L., Talou T. “Examples of semiconductor gas sensors applied to meat products”,

Sci.Aliments 13, 141–148, 1993.

[2] Borjesson T., Eklov T., Jonsson A., Sundgren H., Schnurer J. “Electronic nose for odor

classification of grains, Cereal Chem.” 73, 457–461, 1996.

[3] Bourrounet B., Talou T., Gaset A. “Application of a multi gas-sensor device in the meat industry

for boar-taint detection”, Sensors Actuat. B 26,250–254, 1995.

[4] Gardner J.W. “Detection of vapours and odours from a multi sensor array using pattern

recognition”, B 4, 108–116, 1991.

[5] Gardner J.W., Hines E.L., Wilkinson M. (1990) “The application of artificial neural networks in

an electronic nose”, Meas. Sci. Technol. 1, 446–451, 1991.

[6] Holmberg M., Winquist F., Lundstrom I., Gardner

J.W., Hines E.L. “Identification of paper quality using a hybrid electronic nose”, Sensors Actuators.

B 27, 246–249, 1995.

[7] Pearce T.C., Gardner J.W., Friel S., Barlett P.N., Blair N. “Electronic nose for monitoring the

flavor of beers”, Analyst 118, 371–377, 1993.

[8] Riva M., Benedetti S., Mannino S. “Shelf life of fresh cut vegetables as measured by an electronic

nose: preliminary study”, Ital. J. Food Sci. 13, 201–211, 1993.

[9] Schaller E., Bosset J.O., Escher F., “Practical experience with the “electronic nose” systems for

monitoring the quality of dairy products”, Chimia 53, 98–102, 1999.

[10] Winquist F., Hornsten E.G., Sundgren H., Lundstrom I. “Performance of an electronic nose for

quality estimation of ground meat”, Meat Sci. Technol. 4, 1493–1500, 1993.

[11] Frank Zee and Jack Judy, “MEMS CHEMICAL GAS SENSOR USING A POLYMER-BASED

ARRAY”, Published at Transducers ’99 - The 10th International Conference on Solid-State Sensors

and Actuators on June 7-10, 1999 in Sendai, Japan.

[12] Gardner, J.W.Gardner, Julian W.; Bartlett, Philip N., “Electronic Noses: Principles and

Applications”.