The nose sensor network

A nose sensor network has been installed because of the requirement of con- trolling all the area of the factory. The sensors are metal oxide semiconductor (MOS) type. In some parts of the development of this research, some electronic noses will form part of the system, not being able to classify them as multi- sensor array but a multi-nose array, because the noses are independent with different heating systems and different localizations. This way we have not only different sensor types (all metal oxide variations) but also different locations of the sensors. So the system approach is a hierarchical nose sensor network. All the noses are plugged to a central computer placed in the control room. Cable connections were used instead of wireless for some reasons. One is that this way batteries are not necessary, so the noses can work nonstop indefinitely. Other

is that there are no possibilities of information to be lost due to connection problems or lost packages. But the problem is that to install them, they have to be in fixed positions and also wiring the cable had problems, because it had to be in a safe place to not break or disconnect and to not disturb the workers.

Six noses are installed along the plant. The noses are installed as goes:

Figure 2.10: Location of the sensors in the Elda factory.

• Nose 1: Located inside the control room (where the microwave oven and fridge are), this nose is endowed with a gas sensor type 2600 (general purposes) and a temperature sensor. In this room not only control issues are executed, but also is where the workers eat and store the food. There are fifteen people (three per turn) that don’t eat at the same time. We do not have the times when they eat because they are variable. The sensor is located exactly over the central system near the fridge and the microwave oven (less than two meters). Figure 2.2d shows this nose.

• Nose 2: The second nose is located in the reactor area. The sensors included in this nose are a gas type 2602 (general air quality checker) and temperature. It is located with nose three in the reactors area. In this area the impurities are removed through a process that takes hours. Because of the reactors process, there are traces of oil on all surfaces that can make the nose to saturate and stop working correctly, a fact that must be taken into account since the drift can be very high. Nose is located on top of a beam that holds the entire reactor complex, that way is less affected by the oil. Figure 2.2b shows nose’s position.

• Nose 3: Third nose is also located in the reactor area with the second one. This nose has a gas sensor type 2611 (perfect for detecting gas leakages, which makes it very useful in this part of the factory), a temperature sensor and a radiation one. Being in the center area of the plant and on

top of the reactor area, radiation sensor will receive a lot of light, so it is possible that the radiation sensor saturates, but also can give the system an idea of when the lights are on or off at night time, so is a good approach to simplify the activity control of the plant. Again, this nose can suffer from high drifting because of the oil leakages of the system, that is why it is located over another beam, the same as nose two, but at four meters from it. Figure 2.2b shows nose’s position.

• Nose 4: Located outside of the control room, this was the last nose to be included in the system. From this position, the nose is near the center of the factory, being in a place where it is not near any emblematic place. Having a 2611 sensor (which gives good response from gas leakages) and a temperature sensor, this nose is useful to register a general view of the air in the plant. Also is near a corridor used often by workers to go from one area of the plant to another and to enter in the control room, so the nose can register many different source VOCs.

• Nose 5: This nose is installed in the loading room. The sensors present are a 2611 type (again perfect for leakage detection), a temperature sensor and an humidity one. Being this room where caustic soda and methanol are loaded, the need of detecting a leakage in this room is essential. This was the first nose to include in the system due to its location importance. Here two tanks mix caustic soda and methanol to generate sodium hydroxide and methoxide. The way of introducing the caustic soda is manual and exposed to air, so the nose should register soda loads as they are made. Methanol is injected in the tank by a pipe, so the compound should not be present in the air and, therefore, the nose should not register it. The room is isolated from the rest of the complex, so this nose should not be exposed to the same air flow as the rest of noses. A skylight allows air renewal to clear it and not allowing the nose to become saturated. Figure 2.2a shows nose’s position.

• Nose 6: The last nose is placed in the centrifuge and washing area. It has a 2620 gas sensor (general purpose), another of humidity and a temperature sensor. In this place the biodiesel is centrifuged. Also there is a tap just behind the nose to extract biodiesel samples to analyze them. Every time the tap opens the nose will be able to register it. Also being near the centrifuge, this nose will react to leaks. This area is also covered with oil, so the nose has been put on a beam without contact with any surface. The figure 2.10 shows the position of the various sensors in a rough map of Elda plant.

To prevent waste accumulation in the sensor and circuitry of the nose, a housing is designed to cover the nose (see Figure 2.11). This housing is made by a 3D printer. The housing leaves the gas sensor exposed while the rest is covered by a plastic box, letting some holes for cables and the status lights to be seen. For more information, see appendix 2.