View
941
Download
14
Category
Preview:
Citation preview
ELECTRONIC NOSE P R E S E N T A T I O N B Y :
S A R A T H P R A S A D K . V
N O : 4 5 , S 7 E C E B
Contents
Introduction
Components
Sensors
Pattern recognition
Advantages & limitations
Applications
Future & conclusion
2
INTRODUCTION
• Volatile organic compounds (VOC) are basic to odours
• The detection of chemicals such as industrial gases
and chemical warfare agents is important to human health and safety and
also for industrial purposes.
• But the human nose is not sensitive to all kind of VOC’s
• It is also important to have a detection system for VOC’s which are
hazardeous to human beings.
• Until now, online communication involved only two of our senses, sense of
sight & sense of hearing. A system to transfer or for the difital
reperesenation of the odour is also a matter of concern.
• These all facts lead to the requirement of an electronic detection system for
odours which is “Electronic Nose”
3
WHAT IS E NOSE ?
An electronic nose (e-nose) is a device that identifies the
specific components of an odor and analyzes its chemical
makeup to identify it.
Can be regarded as a modular system comprising a set of active
materials which detect the odour, associated sensors which
transduce the chemical quantity into electrical signals, followed
by appropriate signal conditioning and processing to classify
known odours or identify unknown odours
4
WHY E NOSE ?
E-nose can be sent to detect toxic and otherwise hazardous
situations that humans may wish to avoid.
Human olfactory system may be unstable with respect to mood
or physical condition, but E-nose are not.
Accurate decision making capabilities because of pattern
matching technology
5
Biological Nose Of all the five senses, olfaction uses the largest part of the
brain and is an essential part of our daily lives. Quantifying smells are useful in gas chromatography
Human nose is very sensitive
Subject to fatigue, inconsistencies, adaptation etc.
Smelling toxic gases may involve risk
6
Fig. Conduction route diagram of living being olfactory system
7
The sensor’s response as electrical signals is recorded and
delivered to the signal processing unit and is processed by
using the signal processing unit, whose output is undergone a
pattern recognition and comparison process that prompts
decision making.
8
SENSOR1
SENSOR2
SENSOR3
A/D
DA
TA
P
RE
PR
OC
ES
SIN
G
THE PATTERN RECOGNITION
KNOWLEDGE BASE SYSTEM
A/D
A/D
……
…… …… ……
DECISION MAKING
O
D
O
U
R
1
Array of sensors Fig. Schematic diagram of an electronic nose
Electronic Nose Correspondence of electronic nose with biological nose
9
Biological nose Electronic nose
Lungs Pump
Mucus, Hair, Membrane Inlet Sampling System
Olfactory cells Sensors
Olfactory vesicle Data pre-processing module
Olfactory centre Pattern recognition module
Nerve Impulses Electrical signal
◦ Sample handling system ◦ Generates the headspace of sample to be analyzed
◦ Exposes the odorant to the sensors
◦ Sensing system ◦ Array of different sensors
◦ Each sensor has different sensitivity to different gases
◦ Produces a pattern characteristic of the odour
◦ Signal processing & pattern recognition
10
Sensing system Using array of sensors, each sensor designed to
respond to a specific chemical
Number of unique sensors must be at least as great as the number of chemicals being monitored
Each element measures a different property of the sensed chemical
Each chemical vapor presented to the sensor array produces a signature or pattern characteristic of the vapor
11 Fig. Response of sensor array to different pure chemicals
Sensing system Use of Artificial neural networks (ANN) along with array.
Used to analyze complex data and to recognize patterns, are showing promising results in chemical vapor recognition.
ANN combined with a sensor array Number of detectable chemicals is greater than that of
sensors
Less selective sensors can be used
Less expensive too
Electronic noses incorporating ANNs have been demonstrated in various applications.
12
Electronic nose sensors 1. Conductivity Sensors (a) Metal Oxide Sensor
Oxides of tin, zinc, titanium etc. doped with platinum – Active material
Doped material deposited between two metal contacts over a resistive heating element
Operating temp.: 200°C-400°C
As VOC passes over the active material, resistance changes
Resistance changes in proportion to the concentration of the VOC.
13
(b) Polymer Sensor
Active material is a conducting polymer
e.g. Polypyroles , thiophenes , indoles etc
When exposed, chemicals forms bond with polymers
Bonding may be ionic or covalent
Transfer of electrons along polymer chain is affected, i.e. conductivity changes
Operate at ambient temperature, no heater required
14
Conductivity Sensors
Metal Oxide Sensor Polymer Sensor
Susceptible to poisoning by sulphur compounds present in the odorant mixture
Difficult and time consuming to electro polymerize the active material
Wide availability Susceptible to humidity & can mask the responses of VOC
Relatively low cost, hence widely used
Electronic interface is simple, suitable for portable instruments
15 Fig. Comparison between metal oxide & polymer sensors
2. Piezoelectric Sensors (a) Quartz crystal microbalance (QCM)
Consists of a resonating disk with metal electrodes on each side connected to lead wire
Resonates at a characteristic frequency (10-30 MHz) when excited with an oscillating signal
Polymer coating serves as sensing material
Gas adsorbed at the surface of the polymer increases the mass, reduces resonance frequency
Reduction is inversely proportional to mass adsorbed by the polymer
16
(b) Surface acoustic-wave (SAW)
An ac signal is applied across the input metal transducer
Fingers of this gas sensor creates an acoustic wave that "surfs" the piezoelectric substrate
When the wave reaches the metal fingers of the output transducer, ac voltage is recreated
Voltage is shifted in phase as a result of the distance travelled.
Phase shift depends on the mass & the absorption properties of the sensing polymer layer
SAW devices are less sensitive than QCMs
17
Limitations:
More complex electronics are needed by these sensors than conductivity ones
Resonant frequencies can drift due to the active membrane ageing
Requires frequency detectors
18
Piezoelectric Sensors
3. MOSFET Sensors
Gate is covered by noble metal catalyst, e.g. platinum, palladium, or iridium
Charge applied to the gate leads to current flow from source to drain
VOCs sweeping over the catalyst forms products that alter the sensor's gate charge
Channel conductivity varies
19
MOSFET Sensors
Advantage:
Can be made with IC fabrication processes, batch to batch variation is minimized
Disadvantage :
Reaction products should penetrate the catalytic metal layer in order to influence the charge
Hermetic seal for the chip’s electrical connections in harsh environments
20
4. Optical Sensors
Glass fibre coated on its sides & ends with a thin active material containing fluorescent dyes
Pulse of light from an external source propagates along the fibre
VOCs can alter the polarity of the dyes
Dyes responds by shifting fluorescent spectrum of the light
Simple fabrication- Fluorescent dyes can easily be coupled
21
Signal processing & pattern recognition Main sequential steps:
Pre-processing
Feature extraction
Classification and
Decision making
Data base of the expected odorant should be compiled
22
Signal processing & pattern recognition
Pre-processing Compensates for sensor drift
Compress the transient response of the sensor array
Reduces sample to sample variations
Feature extraction Reduce the dimensionality of the measurement space
Can be more readily inspected visually
Extract information relevant for pattern recognition
Performed with linear transformations e.g. PCA & LDA
Nonlinear transforms, e.g. Sammon nonlinear maps and Kohonen self organizing maps
23
Signal processing & pattern recognition
24
Classification
Bayesian classifiers, Artificial Neural Network(ANN) etc are used
Trained to identify the patterns that are representative of each odour
Identify the odorant by comparing it with trained ones
Decision Making
Used for application specific knowledge
Can determine whether given belong to any one in database or not
Electronic Nose Advantages
Detection of poisonous gas is possible
Can be done in real time for long periods
Cheaper than Trained human sniffers
Individuals vary, e-nose don’t
Digital represenation of odour is possible.
Limitations Time delay between successive tests
Insensitivity to some species
According to application, e-nose has to be changed
25
Applications :
Medical diagnosis and health monitoring
• Can diagnose respiratory infections such as pneumonia
• Examine odour from the body (e.g., breath, wounds,body fluids, etc.) and
identify possible problems.
• Also being studied as a diagnostic tool for lung cancer.
Environmental applications
• Analysis of fuel mixtures ,detection of oil leaks , testing ground water for
odors, and identification of household odors
• Potential applications include identification of toxic wastes, air quality
monitoring, and monitoring factory emissions.
26
Applications :
Military and security applications
• Detection of explosives and hazardous chemicals
• To detect drug odours despite other airborne odours capable of confusing
police dogs
Space applications---e-nose and NASA
• It is a full-time, continuously operating event monitor used in the
International Space Station.
• Designed to detect air contamination from spills and leaks in the crew
habitat
27
Applications :
Food Industry
• Inspection of food, beverage quality by odor,
• Inspection of fish,meat etc
• Monitoring the fermentation , identification of bacteria
• Automated flavor control etc
• Characterizing coffe beans,dates,chocolates.
• To determine trimethylamine (TMA) in milk etc etc
28
In process and production departments
In In quality control laboratories for at line quality control
Future scope :
Research is being done on IC E-Noses
Miniaturizing current Technology
Improvement in sensitivity for lower levels of organisms or smaller samples
Minimizing cost
29
Future scope :
30
Digital Scent Technology
• To sense, transmit & receive smell through internet
• Scent is detected and processed using E Nose
• A scent is indexed along two parameters, its chemical makeup and its place
in the scent ”spectrum” and then digitized into a small file.
• Broadcast: The digital file is sent, attached to enhanced web content.
• Smell Synthesizer: The smell synthesizer means the device which is used
to generate the smells
Conclusion Humans are not well suited for repetitive tasks.
Electronic nose has the potential to become standard tool for smelling. Researches are still going on to make electronic nose much more compact than the present one and to make e-nose ICs.
It also initiates for a new era of representing odour as an electronic/digital data like we have done on picture and sound.
31
32
References
[1] “An Electronic Nose for Multimedia Applications “ by Rafael Castro, Mrinal Kr. Mandal, Member IEEE, Peter Ajemba, and Mujtaba A. Istihad 0098 3063/00 $10.00 © 2003 IEEE.
[2] “Electronic noses and their applications“ by E. Keller1, 0 Approved by Pacific Northwest National Laboratory (PNNL).
[3] “Electronic Nose: Current Status and Future Trends“ by Frank Ro¨ck, Nicolae Barsan, and Udo Weimar Institute of Physical and Theoretical Chemistry, University of Tu¨bingen, Auf der Morgenstelle 15, 72076 Tu¨bingen, Germany Received August 23, 2007
Recommended