More ENose FAQS
SECTION 1 - HOW ELECTRONIC NOSES WORK IN GENERAL AND WHAT THE ENOSE CAN DO
1. Why does an electronic nose use an array of sensors? Why don’t you just make sensors that are specific to the things you are monitoring for?
If you are only looking for 20 things, you only would need 20 sensors. There are very few chemical sensors that are really specific. Most sensors respond to those chemicals they are designed to detect, as well as other chemical species that have a similar chemical structure. The more species you try to detect with specific sensors, the more likely you are to find that sensor responses overlap. Also, if we had 20 truly specific sensors, we would not be able to expand the set of things we are looking for without adding sensors. With an electronic nose, the set of target analytes can be expanded without changing or adding to the sensors.
Using an array of sensors is taking advantage of the fact that most sensors are not truly specific. Each sensor in the array will respond strongly or weakly to each target analyte, and a pattern of response, which we call the fingerprint (see #6), can be created using software. The fingerprint can be identified from known fingerprints which are stored in our data library.
This is, in general, the same way our own noses work. When the air changes and we smell something, the receptors in our noses send a signal to our brains. Our brains create a pattern out of the signals sent by the receptors, where the signal is larger or smaller depending on how strongly the receptor has responded to the odor. The brain takes the pattern and figures out what caused it by looking in our “data library” or memory.
In some ways, the human nose is better than any electronic nose. We can smell a complex odor, such as the odor from a pot of several things cooking on the stove, and pick out the various things put into it. In principle, an electronic nose could do this, too, but it is a complex computational problem, and the computers we have would take quite a while to solve it.
In other ways, the ENose is better than the human nose. For example, we cannot smell mercury, but the ENose can. We become accustomed to an odor, or get tired, or get a cold, and then our noses don’t work as well as we always want them to. We do not really use electronic noses like human noses, rather we use them as monitors to look for the sudden appearance of a limited set of chemicals.
2. What can the ENose really do? Is it like a tricorder on Star Trek? Can you just sweep around a room and tell everything that is in it?
The JPL ENose is an event monitor. It is designed to run all the time and monitor for changes in the air. If there is a change, something could be leaking, something could have been spilled, or there might be some other reason for the change, such as cooking food. The ENose has been designed to detect the presence of a small group of chemicals which could be dangerous if they are present in the air. It is not designed to analyze the air and tell everything in it; that is a job for a different instrument, such as JPL’s Vehicle Cabin Air Monitor (VCAM), a miniaturized combination a Gas Chromatograph and a Mass Spectrometer (GC-MS). GC-MS is a standard instrument for chemical analysis, and is found in analytical chemistry laboratories. The JPL ENose is not meant to replace GC-MS, but to act as a complement to it for air quality monitoring and analysis. In the Space Station, ENose may act as a trigger for VCAM.
3. How does ENose sensitivity compare to the sensitivity of a human nose? A dog nose?
For some things, the ENose is more sensitive than the human nose; for some things it is not. That depends on the group of sensors that have been selected for the array. In addition, the ENose can detect some things that humans and other mammals cannot, such as mercury.
The list of Third Generation Analytes shows the concentrations the ENose will detect and the odor threshold for each analyte. However, the ENose is not able to pick out several different odors at once, nor, at this point, does it have several hundred odors in its data library, as humans do. Dog noses are more sensitive to many chemicals than human noses are.
4. Can you use it to find out what is causing a funny smell or where a gas leak is coming from?
In its current design, the ENose is meant to sit in one place and monitor the air; it is not designed to be moved around or used as a sniffer. In a future version, it could be designed that way, but it would require changing some of the sensors, changing the analysis software, and changing the operational scenario.
5. Is there anything that ENose can't smell?
There are plenty of things the JPL ENose cannot smell as it is now designed; however, there is no fundamental reason the materials in the sensor array cannot be selected to detect almost any chemical species.
SECTION 2 - HOW THE ENOSE WORKS AND WHAT IS IN IT
6. What is an ENose fingerprint?
An ENose fingerprint is a way to display the various responses of the sensors in the array to a particular event. For example, the following “fingerprint” response to ammonia shows that a few sensors (#2, 10, 16, and 23) respond strongly to ammonia, while the others do not. It looks very different from the fingerprint response to acetone, on the left. We can see visually that the patterns are different; in operation the ENose determines this difference using our data analysis algorithm (see #7).
7. Do you use neural networks to analyze your data?
No, we do not use neural networks, because we have found that using a data analysis algorithm based on Levenberg-Marquardt Non-Linear Least Squares Fitting works better for the data we have. NASA wants us to quantify as well as identify the analytes we find, and the data analysis approach we are using gives us the best results.
8. What is special about the ENose polymers, isn't nylon a polymer? How do you make the polymers conductive and why don't you use polymers that are already conductive?
There are thousands of polymers we could use as the basis for ENose sensors. Nylon is one of them. So are polyethylene, which is used to make many plastic products, poly vinyl chloride, which is used to make plastic pipe, and poly vinyl alcohol, which is used to make glue. We select the polymers which we can use to make sensors which respond to our targets, and make them conductive by adding carbon to them.
We could use conductive polymers, such as polyaniline or polypyrrole to make sensors, and there are other electronic noses which do use those polymers. We do not choose to use those polymers because we have found that the approach we use makes sensors that last longer than conductive polymers.
9. What makes the 3rd Generation ENose better than the First?
The 3rd Generation ENose is being made especially for use in the International Space Station. The 3rd Generation ENose is based on the 2nd Generation ENose. For the 2nd Generation ENose, we used the same basic sensor design as in the 1st, but we studied the polymers and how to make them, and have found ways to make them more sensitive than the ones used in the 1st Generation ENose. In the 2nd Generation, we detected chemical species at lower concentration than in the 1st Generation ENose. Also, in the 3rd Generation, we have added the ability to detect mercury and sulfur dioxide.
10. Will faster computers help the ENose work?
Faster computers will help the ENose a lot. In the past 5 years we have added real time data analysis, which can be done only because we have faster computers. In the future, we hope to be able to expand the number of chemical species we can detect and to be able to analyze mixtures of several chemical species.
11. How do you know what the ENose is smelling or how much of it is there when looking at a graph?
We don’t. The graphs we can see on the screen while the ENose is running are only showing us the data as they come in; they are not showing us the results of the analysis. The analysis runs on the computer on the 3rd Generation ENose, and prints the name of the detected analyte and the concentration in a file or on the computer screen.
SECTION 3 - THE ENOSE IN THE SPACE STATION
12. Why do we need an electronic nose in space? What do they have in the Space Shuttle and Space Station now?
NASA would like to have an air quality monitor in the crew areas in spacecraft and in future habitats on other planets. If we spill something like ammonia or alcohol in our houses, we can always open the windows to air it out; astronauts cannot do that. The JPL ENose is an event monitor (see # 1), and is there to warn astronauts if something they cannot smell, but which may be harmful, is in the air. Right now, the air in the Space Station is analyzed by the Volatile Organics Analyzer about once a week, but NASA wants to have something that runs all the time (see # 2). Right now in the Space Shuttle, all the astronauts have to monitor the air is their own noses. There are several experimental devices that are being tested or have been tested in the Space Station; time will tell which ones are selected for full-time use.
13. Why do you have a limited list of targets that you monitor for? Why those particular analytes? Why does NASA care about contamination from ethanol or acetone?
NASA has put together a list of chemical species that could be present in spacecraft air and that would be harmful if they were there. The concentration which NASA has determined to be harmful is the Spacecraft Maximum Allowable Concentration (SMAC), which is set for each possible chemical.
Our list is rather small right now, only 10 species, because we are building the 3rd Generation ENose as an experiment. If it works well during its six month experiment, NASA will decide whether to use it permanently.
Although acetone and ethanol are not very toxic, they can be harmful if people are exposed to them for long periods. Remember, you cannot open the windows to air out the room in the Space Station!
14. How serious is 1 ppm contamination?
How serious 1 part-per-million is depends on the chemical. 1 ppm of acetone is not serious - you are exposed to much more than that when you use nail polish or nail polish remover. 1 ppm of formaldehyde is, however, much more serious, and could cause headaches and eye irritation in the short term and permanent damage in the long term. The 30 day SMAC for ammonia is 1.5 ppm; it could cause liver damage if a person were exposed to a few ppm of ammonia continuously for 30 days. Most people cannot smell ammonia until it reaches 30 - 50 ppm.
SECTION 4 - THE ENOSE PAST AND FUTURE
15. When was the ENose invented?
The concept of an electronic nose as we understand it now was first reported in 1982 by K. Persaud and G. Dodd at the University of Warwick in Great Britain. Several researchers have worked on electronic noses since then, using several different types of sensors.
At JPL, we started working on our ENose in 1995. The 1st Generation JPL ENose, which flew on Space Shuttle STS-95 in 1998, was the first electronic nose which quantified as well as identified target analytes.
16. When will the ENose be available for commercial applications? How much does one cost and where can I buy one?
JPL will not be making ENoses for consumer use; we are working specifically toward the NASA use in crew habitat in space. However, we have considered several other possible uses for the JPL ENose, and are doing research into how to modify our ENose for those applications. If they are successful, it will be up to a company to decide to make ENoses.
There are already several companies that make electronic noses for various applications. You can buy them from these companies; if you are really interested, search on “electronic noses” on the internet, and see what you can find.
17. Can the ENose be used to detect explosives or harmful chemicals released by terrorists?
In principle, yes, the JPL ENose, or many other electronic noses, could be used in homeland security types of applications. We are not working toward that at JPL, but many other researchers in other laboratories are.
18. Can it be made smaller / lower power? Could it be made really small, like a human nose?
Yes, making the ENose smaller is primarily a matter of designing the electronics to be smaller. How small it can actually get will depend very much on how you want to use it.
We get air to the receptors in our noses by breathing. The ENose also has to have a way to get air to the sensors. Right now we use a pump, so the size of the ENose is somewhat limited by the size of pump we can get. In another application, however, such as one where the device moves through the air, the air could be forced into the sensors by the movement. In that case, the ENose could be very small.
19. How frequently does the ENose record the sensor resistance?
We record resistance one time every 20 seconds, or 3 times a minute. We chose this frequency because it gives us a good picture of the air without filling our computer memory too fast. There are many other ways this could be done. For example, we could take one point every minute, then when we see a change, take points more frequently until we understand the change. This approach could save power or make a battery powered system run much longer.
20. How does the ENose get its data library?
We “train” the ENose in the lab by exposing the sensors to the target analytes under various conditions. Data from that training is used to construct the data library.
We also have developed a theoretical model of how sensors will respond to various types of chemicals, so we can calculate how the ENose will respond, or, if we have a signal from something new, we can calculate what may have caused it. Unfortunately, nothing works like it does on TV or in the movies. Our technology is good, and our computers are good, but we do not have access to magic.