Show page

Differences

This shows you the differences between two versions of the page.

--- iothings:proiecte:2021:iotairquality [2022/01/27 19:29]
sabin.olaru
+++ iothings:proiecte:2021:iotairquality [2022/01/27 20:29] (current)
sabin.olaru
@@ Line 91: / Line 91: @@
 == 4. Results ==
+==Outdoor measurements==
+The measurements were carried out outdoors, placing the sensor box outside the window of their own house, located on the fifth floor of a ten-story building. An important aspect to note is that the part of the block from which the measurements were made is not oriented towards the boulevard, so the impact of traffic on the measured air quality is lower. had a duration of seven weeks, being performed daily in a number of almost 7000 measurements, totaling a number of 47016 measurements. temperature and humidity:
+During the sampling period, there were precipitations, atmospheric phenomena which are reflected in the high percentage of measured humidity. The measured daily average concentrations of the pollutants and shown in the following table:
+{{https://ocw.cs.pub.ro/courses/_media/iothings/proiecte/2021/sabintabelmasuratori.png?450x250}}
+Based on the measured concentrations, the AQI indices were calculated for each monitored pollutant, and then the maxima were designated as the total AQI indices measured. The values of these indices are listed in the following table.
+{{https://ocw.cs.pub.ro/courses/_media/iothings/proiecte/2021/sabintabelaqi.png?450x250}}
+These measurements show a phenomenon that was to be expected in the environment in which the measurements were performed, namely that the main pollutant and the one that gives value of the total AQI mostly is PM2.5 (i.e dust particles in suspension). In the open air, in an urban environment, the phenomenon was to be expected, being a proven subject and brought into discussion by numerous scientific researches. It can also be seen that on Wednesday, the AQI calculated on the basis of the concentration of PM2.5 is significantly lower than in the other days. This result is marked in the table with the symbol * to highlight its particularity. This observation led to the verification of the sampled data and an interval was identified during which the sensor did not work value 0. By removing the erroneous measurements, a new AQI value corresponding to PM2.5 of 42.69 was obtained. This newly obtained value was also considered as the value of the total AQI for the day in question.
+== Comparison to existing models ==
+In order to compare the performance of the designed system with existing models, two reference sources were used for the air quality index:
+  * **waqi.info** - An online source that monitors air quality in more than 100 Countries with over 30,000 measuring states in 1,000 major cities, which base their AQI calculation on the following pollutants: PM2.5 particulate matter and PM10, ozone, nitrogen dioxide, sulfur dioxide and carbon monoxide.
+  * **iqair.com** - A source available online of Swiss origin, which provides information on global air quality, and which provides data from four locations located in Bucharest.
+The process of taking data from each source was made differently, due to the different distribution of their measurement stations. Source waqi.info has multiple stations nearby the location where our own measurements were made, but none close enough to be chosen to the detriment of others. Thus, the solution was chosen to take the data from the nearby stations and to average them, obtaining an average index that will be later compared to the data obtained by its own system. The second source, iqair.com, provides a measurement point located close to the location where the measurements were made, and thus being chosen as a reference.
+The data collected from these references were collected daily and tried to be as frequent as possible
+{{https://ocw.cs.pub.ro/courses/_media/iothings/proiecte/2021/sabingraficaqi.png?450x300}}
 == 5. Conclusions and perspectives for further development ==
 == Evaluation of the parameters of interest ==
@@ Line 104: / Line 128: @@
 == Prospects for improvement and further development ==
 ==Improvements==
-Although the performance of the realized system is very good, and the parameters of interest have been mainly fulfilled and successfully, there are improvements, which can or could be improved, and even more the quality of this system.
+Although the performance of the realized system is very good, and the parameters of interest have been mainly fulfilled and successfully, there are improvements which could be done to improve the quality of this system even more.
 ==Measurement system power supply==
-The external battery model that was used in this work is an older model that has been used before, so its capacity is no longer at the original level. The reduction of the system autonomy, as well as the decrease of its weight, the model used having a considerably higher weight compared to the new models.
+The external battery model that was used in this work is an older model that has been used before, so its capacity is no longer at the original level. Replacing it would lead to an increase of the system autonomy, as well as the decrease of its weight, the model used having a considerably higher weight compared to the new models.
 An additional solution for the improvement of the system autonomy would be using an alternative method of continuous operation of the sensors, for example turning them on for shorter periods of time in which to perform measurements, and then turning them off at the next predetermined measurement interval.
-This improvement proposal, however, needs to be carefully evaluated, preferably through experiments, because the sensors, such as the MQ gas, may need a period of prior operation before it can provide data. to warm up and calibrate. Thus, although autonomy would be improved, the quality of the data provided by the sensors could suffer.
+This improvement proposal, however, needs to be carefully evaluated, preferably through experiments, because the sensors, such as the MQ gas, may need a period of prior operation before it can provide meaningful data, as they need to warm up and calibrate. Thus, although autonomy would be improved, the quality of the data provided by the sensors could suffer.
 ==Measured range of indicators==
-The built-up system does not currently cover all the polluting agents of interest, and in this respect it is very suitable. Although the performance of the system was good, the addition of additional sensors would provide a more comprehensive and detailed picture of the air quality. It should be noted, however, that the addition of new sensors will have the effect of reducing the autonomy and possibly the need to add another breadboard, the current one being used close to its maximum capacity.
+The built-up system does not currently cover all the polluting agents of interest. Although the performance of the system was good, the addition of additional sensors would provide a more comprehensive and detailed picture of the air quality. It should be noted, however, that the addition of new sensors will have the effect of reducing the autonomy and possibly the need to add another breadboard, the current one being used close to its maximum capacity.
+==Accuracy of data provided by the system==
+Although the comparison with the existing models showed a good performance of the system regarding the accuracy of the measured data, their calibration, especially in the case of MQ sensors was performed empirically.
+A more accurate method involves the use of professional apparatus, but the costs of these are high, so that the improvement considered for the development of the work is the use of machine learning techniques.
+To determine the formulas and correction factors that provide the most accurate data. It has also been pointed out that a correction may also be required in the case of the temperature sensor to cancel out the effect the heat dissipated by the system has on the measurements.
+==Further development==
+The desired direction of development of this work is the realization of the improvements mentioned above and the realization of a network formed by systems identical to the built one. For geographical distribution, the installation of each system on a drone is considered , with the weight and the dimensions of the system allowing the implementation of this concept.
+Based on the measurements, we could then make a three-dimensional representation of the air quality, available in the Web application. On the basis of that, we can then analyze the quality and evolution of the quality of the air in an area.
+In case of need, the assembly could be relocated to any area of interest, taking into account the portability of the measurement system and the mobility of drones.
+== 6. Bibliography ==
+[[https://www.who.int/health-topics/air-pollution/]]
+[[https://medium.com/datadriveninvestor/4-stages-of-iot-architecture-explained-in-simple-words-b2ea8b4f777/]]
+[[https://www.gsma.com/iot/wp-content/uploads/2018/08/Air-Quality-Monitoring-ReportIoT-Big-Data0818.pdf]]
+[[Perry Lea.”IoT and Edge Computing for Architects - Second Edition”.  2020]]
+[[David Hanes Rob Barton, Gonzalo Salgueiro.”IoT Fundamentals:  Networking Technologies, Protocols, and Use Cases for the Internet of Things”.  2017.]]
+[[https://www.eea.europa.eu/themes/air/air-quality-index]]
+== 7. Demo==