As cities keep growing in size, so does pollution and health hazards threatening citizen's lives. Monitoring these hazards is becoming ever more complex and there is a need for a new system capable of capturing this large amount of information. For instance, noise pollution increases unstoppably in cities but the monitoring of this pollution doesn’t meet the same growth. How can an interconnected IoT sensor device monitor a parameter such as noise pollution over a determined zone?
The main idea is the creation of the “Home Analysis Station” (HAS). This device is a friendly-to-use measuring tool that individuals with no experience could use to ressource data about measurable parameters like air pollution(inside and/or outside), the quality of a sample of water, or even noise pollution. The HAS has two principal advantages: it can give data about the quality of life within the household owning the system but it can also be used cooperatively to approximate the data of a given zone. There are two main obstacles towards the realization of the HAS: making it accessible to virtually every household and creating a network so that the interested institutions can gather data about a large zone
The former determines in great part the structure of the HAS. It must be an easy to use, cheap and compact system with a friendly interface. Owning a HAS should be pretty similar to having a station like ¨Alexa¨ from Amazon or Google Home; smartphone connectivity is a must. A Raspberry PI takes care of the information chain, receiving the data and sharing in to the internet. Sensors can be implemented in an almost modular fashion, giving HAS an incredible potential. Creating the network consists less on the system itself and more on collaborative work. To do so, many incentives could be employed such as tax deduction or monetary rewards for the people sharing their data, respectively if it is governmental or private institutions who are interested. Then, gathering data from specific houses distributed across the area, this discretization gives an accurate estimate of the state of the measured parameters on the zone. This is why it is important for HAS to be as accessible to everyone as possible, since it needs to be evenly distributed and making it too hard to use would mean only certain individuals would be interested in learning how to use it.
Although the HAS is pretty modular in nature and has great potential as a measuring tool of countless parameters, this project will focus on the analysis of noise pollution, leaving water and air quality analysis as possible future upgrades. Capturing sound waves requires a microphone to be either present on the HAS or connected from a distance. However, measuring raw noise intensity can be misleading because noise pollution isn’t strictly any sound, nature and distance of the source must be factored in to get accurate measures. Instead, the HAS collects audio files and makes an analysis through a “machine learning” algorithm to associate certain frequencies and their harmonics to noise coming from traffic or other important sources of noise pollution. This way the intensity and more specifically the variation on intensity of these specific portions relate more accurately to actual noise pollution.
Overall, the HAS is a surprisingly simple system, and has immense potential because of that simplicity and because of its potential. What is left is leading HAS towards a more complete analysis station, starting for instance with an indoor air quality monitoring, helping in the detection of Carbon Monoxide (CO) or CO2 rates, especially in houses for the elderly, where these factors can be dangerously neglected.