Although the least understood human sense is that of smell, it is perhaps the most interesting one. Indeed, the human nose is used routinely as an analytical tool to assess the quality of foodstuffs, drinks, perfumes, and many other household products. Yet its practical application is severely limited by the fact that our sense of smell is subjective, tires easily, and is therefore both expensive and difficult to use. Consequently, there is a considerable need for an instrument that can mimic the human sense of smell and be used in routine applications.
Since the mid-1980s, there has been increasing interest in developing so-called "electronic noses" (e-noses), that is, electronic instruments that can detect and recognize simple and complex odors.1 And since the mid-1990s, nearly 20 years after the concept was originally published, the commercialization of e-noses has started to take place. The main reasons for this delay were the complex nature of the problem and the need for advanced technologies. However, the recent development of microsensor technology has led to low-cost integrated chemical sensors and application-specific microprocessing devices. This, coupled with our greater understanding of artificial intelligence, has allowed us to construct electronic instruments that perform in a manner similar to our own olfactory system.
E-noses are now being developed as systems for the automated detection and classification of odors, vapors, and gases. They are generally composed of a chemical sensing system (e.g., sensor array or spectrometer) and a pattern recognition system, such as an artificial neural network (ANN). At Pacific Northwest National Laboratory (PNNL), e-noses use ANN technology for the automated identification of volatile chemicals used in environmental and medical applications.2
The electronic nose works as follows. While a chemical vapor or odor is blown over a sensor array, sensor signals are digitized and fed into a computer. The ANN (implemented in software) then identifies the chemical. The benefits of e-noses include compactness, portability, real-time analysis, and automation.
Currently, the largest market for e-noses is the food industry. In some instances, e-noses can augment or replace panels of human experts and can reduce the amount of analytical chemistry performed in food production, especially when only qualitative results will do.
An electronic smelling device is a valuable tool for analyzing whether a product has gone bad. Potential applications of e-noses in the food industry are numerous: inspecting and grading food quality by odor; inspecting fish and beverage containers; controlling fermentation, automated flavoring, and microwave cooking; monitoring the ripening of cheese; verifying if orange juice is natural and/or fresh; testing plastic wrap for containing the odor of onions; and classifying grains and blueberry ripeness.
Using human odor panels to evaluate and control the quality of raw materials or finished products is extremely labor intensive, time consuming, expensive, and error prone. E-noses can quickly identify a characteristic odor classified as "good" or "bad" by the odor panel, thereby decreasing the workload, improving throughput, and reducing the cost of screening many samples at different stages of the manufacturing process. The system is applied easily to the manufacture and quality control of perfumes, cosmetics, and fine chemicals, as well as to packaging, monitoring environmental quality, the automotive industry, medical and diagnostic matters, and microbial classification.
The PNNL is exploring the technologies required to perform cost-effective environmental restoration and waste management. This effort includes developing portable, inexpensive systems that can identify contaminants in the field in real time. Environmental applications of e-noses include identifying toxic wastes and household odors; analyzing fuel mixtures; detecting oil leaks; monitoring air quality, factory emissions, and hazardous chemicals; and testing ground water for odors.
Since the sense of smell is important for physicians, an e-nose can be used as a diagnostic tool to examine bodily odors (e.g., breath, wounds, bodily fluids, etc.) and identify possible problems. Odors in the breath can indicate gastrointestinal, sinus, and liver problems, as well as infections and diabetes. Infected wounds and tissues emit distinctive odors, and odors coming from such bodily fluids as blood and urine can indicate liver and bladder problems. Currently, an e-nose for examining wound-related infections is being tested at South Manchester University Hospital.
In similar applications, ANNs have been used to track glucose levels in diabetics, determine ion levels in bodily fluids, and detect such pathological conditions as tuberculosis.
While the inclusion of visual, aural, and tactile senses into telepresent systems is widespread, the sense of smell has been largely ignored. PNNL recently proposed a more futuristic application of e-noses for telesurgery. In this application, an e-nose would identify odors in a remote surgical environment. These identified odors then would be transmitted electronically to another site, where an odor generation system would recreate them.
The next decade should see the cost of e-noses fall dramatically, with the result that they will be used not only in industry but also in everyday life. They can be used, for example, to detect tainted foods in the refrigerator, ensure clean clothes in the washing machine, detect poor air quality in the car, and perhaps even help us monitor our own health.