| Introduction | | | | 1. Static odour delivery. |
| Electronic Nose (eNose) is a device used to detect | | | | 2. Mass-flow systems. |
| and recognize odours/vapours, i.e. a machine olfaction | | | | As the two names suggest, the basic mechanism for |
| device with an array of chemical sensors. | | | | the first type is that there is no odour flow but |
| Alternatively, according to Gardner and Bartlett, | | | | simply a flask contains the sensors array with a fan |
| (1994) definition [3]: | | | | at the top to distribute the flow within the flask. This |
| "An electronic nose is an instrument which comprises | | | | type was the design of the first eNose in 1982. |
| an array of electronic chemical sensors with partial | | | | The second type which is very popular now is where |
| specificity and an appropriate pattern recognition | | | | the odour flows within the system. Most eNoses |
| system, capable of recognizing simple or complex | | | | designs are made in this way. |
| odours" | | | | To complete this brief historical outlook concerning |
| The most common use at the present time for the | | | | the eNose, it is a good idea to look at the basic |
| eNose is within the food and drink industries. In | | | | schematic comparison between human and electronic |
| addition to this field, eNose can be used in other | | | | noses [6], summarized in the following two sections. |
| areas such as petroleum qualitative and quantitative | | | | The Human Nose[6] |
| analysis, detection of explosives, classification and | | | | There are millions of self generated receptors (over |
| degradation studies of olive oils, development of a | | | | 100 million) with selectivity classes can range from 10 |
| field odour detector for environmental applications, | | | | to 100. |
| quality control applications in the automotive industry, | | | | The human nose is very adaptive but unlike the |
| discrimination between clean and contaminated cows' | | | | eNose, saturation can happen and that is one of the |
| teats in a milking system, cosmetic raw materials | | | | reasons why it operate only for a short periods of |
| analysis, plus many other important areas such as in | | | | time. Variety of odours can be identified, plus it can |
| the medical and space fields. | | | | detect some specific molecules but it cannot detect |
| The principle of eNose is that it uses an array of | | | | some other types of simpler molecules. |
| sensors, whether in the form of different types of | | | | As a biological system, infection can take place, which |
| polymers or via the use of metal oxide semi | | | | may affect the ability to smell. |
| conductors, the principle here is still the same. | | | | And finally, smelling can be associated with various |
| When molecules from any element deposited on the | | | | experience and memory. |
| surface of the sensor, the electrical conductivity | | | | The eNose [6] |
| changes, as and when the surface expands. This is | | | | Approximately 5 - 100 chemical sensors manually |
| the basic idea of how eNose works i.e. change of the | | | | replaced. In comparison with the human nose, it is not |
| sensor resistance when the sensor exposed to | | | | possible to reduce automatically the number of signals |
| odours/vapours. | | | | to a particular one. |
| The pattern displayed on the monitor for each | | | | As the eNose continue to develop, it is possible in the |
| particular resistance is unique (i.e. the type of odour | | | | future to become adaptive, it is also unlikely to |
| or vapour of a particular sample). In this way it is | | | | become saturated and can work for a long periods |
| possible to distinguish a sample from another or the | | | | of time. |
| state/condition of the sample itself, as the headspace | | | | If pattern recognition hardware provided within the |
| from each sample has a unique signature on the | | | | device, then new real-time signal treatment can |
| eNose sensors resistance. | | | | occur. Unlike the human nose, eNose needs to be |
| Brief History | | | | trained for each application. It can detect simple |
| It is difficult to pin point the exact date of "when and | | | | molecules but it cannot detect some complex |
| how" the idea of designing a system, which can | | | | molecules at a low concentration. |
| mimic the human nose, came about. However, the | | | | The eNose can get poisoned (sensors' malfunction); |
| following dates with devices give a better | | | | at the same time it is possible for eNose with |
| understanding of how the design progressed for a | | | | multi-sensors to be associated with other functions |
| machine olfaction devices (MOD) system. The MOD | | | | and recognitions. |
| design led eventually for the conceptualisation of the | | | | How eNose work? |
| eNose. | | | | A number of operation parameters are usually |
| Please note that an eNose differ from other types | | | | required in order for the eNose being able to function |
| of MOD by simply having multiples sensors, while | | | | "to a maximum effect". These operation parameters |
| other devices may have one sensor only or simply | | | | can be: |
| the mechanism itself differ substantially from the | | | | |
| eNose basic working principles. | | | | 1. Setting up the temperature for the sample |
| The name MOD, therefore, cover devices such as | | | | incubation |
| eNoses i.e. devices with multiple sensors, as well as | | | | 2. The size of the sample. |
| devices with single sensors - or those devices which | | | | 3. The rate of injection. |
| operate on a different design principles. | | | | 4. The quantity of injection. |
| The four following dates are important in the history | | | | 5. The added solvent being used. |
| and development of the eNose: | | | | 6. Flow rate. |
| 1. The making of the first gas sensor, Hartman 1954 | | | | 7. Sensor type. |
| 2. Constructing array of 6 termistors, Moncrief 1961 | | | | 8. Sensor operational parameters. |
| 3. First Electronic Nose, Persaud and Dodd, 1982 | | | | The above are just examples; however, there can |
| 4. Ikegami (Hitachi Research Laboratory, J) array for | | | | be other factors as well. |
| odour quality - 1985 | | | | As mentioned briefly earlier on, the principle of eNose |
| Therefore, the first recorded scientific attempt to | | | | is mainly rests with the one or more (an array) of |
| use sensor arrays to emulate and understand | | | | vapour-sensitive detectors (sensors). Usually the |
| mammalian olfaction was carried out by Persaud and | | | | detector is made up from certain type of sensitive |
| Dodd in 1982 [3], at the University of Manchester | | | | materials which its characteristic or behaviour change |
| Institute of Science and Technology. | | | | in response to absorbed or adsorbed molecules. As |
| A device was built with an array of three metal-oxide | | | | we measure the changes in each sensor, identification |
| gas sensors used to discriminate among twenty | | | | can be made for the unknown odour(s) by |
| odorous substances. Using visual comparison for the | | | | comparing it with the library data. |
| ratios of the sensor responses, they obtained the | | | | ConclusioneNose devices have been developed over |
| pattern classification. | | | | the past 20 years to perform a variety of |
| The name itself "Electronic Nose" used for the first | | | | identification tasks in various industries. However, |
| time during 1988 and has come into common usage | | | | merely a few years ago, the majority of work and |
| "as a generic term for an array of chemical gas | | | | publication related to this field were mostly restricted |
| sensors incorporated into an artificial olfaction device" | | | | to the area of research. These days, various types |
| [3][4] after the introduction of this title at a | | | | of commercially available eNoses can be purchased |
| conference covering this field in Iceland 1991. From | | | | anywhere in the world. |
| that point, the idea and the principles of the eNose | | | | The reason for the relatively fast development and |
| has grown and developed into different fields across | | | | commercialization of these devices is because they |
| the globe. | | | | attracted new interest in their application in the fields |
| Historically speaking, there are two different types of | | | | of food, environment, medical diagnosis, industries, |
| eNoses (Pearce 1997): | | | | security and other related areas. |
| | | | |