We cannot erase the air & sound pollution in a snap of the fingers, but we can use equipment to hedge it. Therefore, we thought about collecting related data in a proficient way and use it to make citizens' life healthier. We plan to map city areas with noise and air sensors in order to provide the healthiest path to runners, walkers, etc. Figure 1 explains the services cover by our solution while the figure 2 explains the cost and the pricing. In the following, we explain how we designed our solution and why it should win.

We chose to build our own LoRa sensor from scratch thanks to a Lora library on Arduino Uno. Thus the implementation will be faster and cheaper. We chose to use the RFM95 Lora module, compatible with Arduino Uno, which is used in the case of monitoring. There are different air pollution indexes (AQI for US, CAQI for Europe, API for Malaysia). We chose 6 kinds of sensors that could help us to evaluate the air pollution index through 6 gases which are CO, O3, PM2.5, PM10, NO2, and SO2, with respectively MQ-9, MQ-131, DSM501A, GP2Y1010AUF, MICS-2714 and 3SP_SO2_20 sensors.

For each measure, it will transmit for each gas: concentration, air pollution index the total air pollution index. Sound pollution in outdoor requires a sensor that can detect a large range of sound, from 30 to 130dB. We chose the Gravity analog sound level meter because it is a sensor module specified for outdoor sound measurements with a high precision of measures. About the physical protection of the device (which is exposed in an urban environment), we chose to design a custom housing with a 3D printer.

For the physical layer, we chose to use LoRa which is an LPWAN technology well suited for monitoring. It has a high link budget of 154dB which assures LoRa indoor and deep indoor penetration. LoRa in an urban city should be able to cover 1Km^2 even taking into consideration all radio effects (shadowing, fading, ...). The power used will be minimized as the area is relatively small in comparison with the LoRa coverage. This is due to the use of the lowest modulation scheme (SF7 or SF6 according to region) for short range coverage. We can thus ensure a device battery life between 5 and 10 years.

LoRa gateway can serve a maximum of 1000 objects. We chose to use 50 to not overload the station and avoid interferences of devices between each other. Besides, we don’t need to pay the frequency we will use, as LoRa is on an unlicensed band. We only need to pay the LoRa module for transmitting data. However, all device should respect a duty cycle of the band, which limits the number of messages we can transmit. Assuming the lowest power modulation scheme (SF7) and maximum payload size, on 868 MHz with a duty cycle of 1%, you need to wait 10 seconds before transmitting again. You can transmit at most around 8500 messages a day. The same reasoning can be done with other frequencies.

For the upper layers, we chose to use the open standard LoRaWAN network which is based on LoRa. We can create our own private network based on open-source code. Therefore only the data of our devices will be transmitted through our gateway instead of using a public LoRaWAN network as TTN’s one where you need to deal with other’s data. It also allows us to cut from network cost operator for LoRaWAN network.

Using LoRaWAN, the network architecture is star topology (cf figure3). We can place the gateway on a rooftop so that it can avoid some interferences from the building itself and we can connect it to ethernet easily. We also chose an omnidirectional antenna in order to reach the same coverage in every direction. Sensors can thus be uniformly distributed in the area we cover. We chose to put at the same place one air pollution sensor and one sound sensor.

About the back-end, we chose Azure because is affordable and has several projects powered with Lora. We will use 50 air pollution sensors and 50 sound pollution sensors to cover 1km^2 area. Using the free Azure plan, with a limitation of 8000 messages per day, each device can send 160 messages per day, which correspond to one message every 9 minutes. We provide a well-documented API to fetch the data, which make super easy to build applications on top of it.

Our solution brings reporting and monitoring of pollution and noise across wide urban areas. These high-quality data are used for an application that can help every person to prevent himself from air and sound pollution. It can also be useful for actors improving citizen’s life with smart city projects. The second illustration highlights the services covered by our solution.

We have an optimal configuration to use LoRa, which provides long battery life and our solution permits high-frequency measurements with large coverage and is easily scalable. 


Although Paris is not a top 10 most polluted city in the world, it happens to have high pollution days, and constant noise areas. Both of us really like to run, but sometimes we can’t because of pollution. When we heard about this challenge, we instantly thought about the opportunity of helping the life of everyone who wants to enjoy the urban life as much as still staying healthy. We then thought of a map helping runners to choose the healthier path. We also realize we can extend the idea to every people in the city because a lot of them are walking to reach public transports or to go directly to their job. Building the map idea, we thought about the architecture and then we got the second idea: using the collected data to provide another service. We have a huge mapping coverage solution with LoraWAN and processing data on server side to assure quality. These are the two mandatory keys to start opening the API. Our main motivation for doing this project is coming from our passion for IoT. We are two students working in IT, one in IoT and the other one in Cybersecurity. We strongly believe in LoRa technology and in particular LoRaWAN which is open source. People sharing knowledge can help to build a better future.


Voting is closed!