 Good morning everyone. My name is Karina. I'm a researcher at IMAG-SMIT-VUB. SMIT stands for Studies in Media and Information Technologies. We are linked to the free University of Brussels and also IMAG. My background is in Social Sciences, meaning that I look to the connection between society and new digital technologies. I will introduce you to the Hack Air Project, which is already running since 2016 and which some of you will already know. The main objective of the project is enabling citizens or communities of citizens to easily set up air quality monitoring networks for measuring outdoor air pollution. We are also focusing on PM, PM2.5 or PM10. The main outcomes that we are envisioning with our project is first of all to have easier access to air quality information. As you know, the information now is quite dispersed or not easy understandable for a common citizen. The second outcome that we are envisioning is to set up a participatory sensing approach. This means actually that we believe in the power of citizens and that we distribute or support them with assembling low-cost sensors to measure outdoor air pollution. Although we know that they are of lower quality than official air quality measurement stations, we do believe that it is important to give them an indication to which we are exposed to in our daily life. And then the second the third outcome is to create awareness and behavior change. Behavior change on different levels, first of all, knowledge. We would like to distribute information about the daily air quality levels and then also how to take action yourself to contribute towards a cleaner air quality or to protect yourself against it. Then I also of course need to mention that this is a European funded project by the European Commission and that everything is provided as open source, both the hardware as the data that is collected. These are the partners that are involved in the project. We have a Greek coordinator called Draxis. They are also responsible for the technical development of the mobile app and the web platform. They are working closely together with CERT, which is responsible for data discovery integration. Then we have a very nice partner from the Netherlands on subject who is helping us with the dissemination exploitation. Then we have two official pilots right now, one in Germany, represented by Bund or Friends of the Earth. They have a very large network of citizens around 400,000 spread all over Germany who are going to be involved in the project. And the other pilot is in Norway, represented by NILU, that is the Norwegian Institute for Air Quality Investigations. They are also representing a network of asthma patients. Beside that, we have three affiliated organizations of which one, Kravis, is here, located in Brussels and his name is Dimitris and he will be coordinating the contacts here in Brussels to form a network around air quality measurements. This is seen as a test case, let's see. Then the concept of HackAir. As you can see in the picture, we have four different ways of collecting data around air quality. First of all, we have mobile images, which I will explain later on. Then we have three different open hardware sensors, low-tech measurement through a cardboard sensor. This is something for kids and youngsters. And then we have open air quality data sets. As you can see, the web platform and the mobile app are currently showing the data from Luftdaten but also from open air queue. So all these different data sources are coming together on one map. Currently, we are supporting three different low-cost sensors. We label them as a hacker mobile and a hacker home. A hacker home, we have two different types here. We are actually using the same PM sensor as Stuttgart is using or as you will assemble this afternoon. But we opted for two different microcontrollers as we were seeing that for non-technical-skilled citizens it was too difficult to assemble the Arduino-based sensor. That is why we opted for Wemos, that is less soldering and putting the pieces together in a more efficient way and it also costs less. Then the hacker mobile is actually targeted towards more expert users and professionals. It is a PSOC-based one. It costs around 50 euros. And that one is something that you can carry along in the city to measure your individual exposure real-time. It works via Bluetooth that sends the data through the mobile application. Then we have the cardboard sensor. So I said this is something for kids or for youngsters that they can do in school to have an indication about air quality. Therefore, you can use a drinks carton from which you cut out a piece five by five. You put some petroleum jelly on it and then you hang it outside for 24 hours. After that, you take a picture of it with a micro lens. So that's something that you need to buy. Then you take a picture of it with the mobile app of Hacker. The mobile app of Hacker has a special algorithm right now to give you an indication about air pollution by just taking an image of the sky or from that particular cardboard sensor. To take a picture of the sky, some conditions are necessary. You need to have a good deal of sky. It may not be too cloudy and not in the early morning or early evening. And then through that specific algorithm that measures light intensities, you will give an indication. Based on that indication, you know how good or how bad the air quality is. This is a screenshot of the web platform that we are currently having. On the right side, there are some filters related to open data. There you can find the data from Luftat and from open air queue. On the left-hand side, there is the index provided that comes from the European Environmental Agency. Related to that, we are also providing tips of the day and personalized recommendations. What you should do or not do when the air quality is bad. So, limited sporting, limited picnicking, for instance, or for asthma patients, we recommend to stay inside, for instance, when the air quality is bad. So then, related to the timeline, as I mentioned before, it's already started in 2016. In that year, we dedicated a lot of activities towards the co-creation of the actual tools. We organized multiple workshops to gather user wants and needs. Then in the second year, everything was developed and arranged together for the sensor parts. And now in the third year, we are engaging the actual communities. And the first week of February, everything was officially done, let's say. Some first results, since the first week of February, is that we see already now that we have 122 active sensors. This is mostly in Germany. The community there is quite interested in installing those sensors. And they also get pre-assembled ones from their organization. They can have the option. They can buy the pre-assembled one, or they can be involved in a workshop where they learned to do it themselves. We see that we have around 480 new user accounts and 465 downloads of the application. Right now, we are having 206 images of the sky that are being uploaded. An increase in engagement or impressions through social media. Now, to conclude what I wanted to focus upon in particular, and what is my activity, especially in the project, is related to engaging communities, because this is a question that is often posed. How do we involve citizens in citizen science, let's say. To start with that, I wanted to show you this framework from Hackley that describes four different levels of citizens' engagement that you can reach. Going from the level of crowdsourcing where citizens are just data collectors or sensors. Here we can actually give the example of Korean users. They are pure data collectors. Level two distributed intelligence. Here, citizens are also going to help in analyzing the data. An example here is, for instance, a galaxy zoo where they are also doing some validation with the pictures. And then we can say that Hackare is positioned in between level three and four. So they are, citizens are being engaged in data collection, data analysis, but also throughout multiple stages in the project. And hereby I'm referring to the co-creation of the solution and the evaluation of it. We do see, however, that citizens still need the help of a scientist to help them interpreting the data. So we are not there yet in terms of literacy skills. Then I also wanted to point out that related to the levels of difficulty, you need to take this into account when you are engaging citizens. So we actually are foreseeing a learning curve for citizens being engaged in the project. Going from a simple task to a more advanced task. Simple being you are submitting your air quality perception. How are you feeling the air today? Or uploading an image of the sky. Going from assembling the sensors. Our engagement strategy or also how we often call it is an engagement-related behavior change approach, as we believe, by being engaged in our project. We are also going to establish some behavior change as final outcome. Herefor we are relying on community-based social marketing and the 7E framework. And the 7E framework looks like this. We actually come up with several different types of behavioral change interventions. And we are using these different tactics to involve citizens in different ways, going from initial participation to continued participation. Three, I wanted to highlight this, for instance, encouraging citizens. This is pointing out to extrinsic motivations through gamification. So by participating in HACA, you are collecting points, you can earn batches. The more you contribute, the more points you gain. Although we see that this was not one of the favorite approaches of the users right now, it's still being involved as a hypothesis if it can lead to behavior change or not. Then, Enlighten. This is a very important one. This is about providing information. We must say that when citizens are involved in citizen science, you also have to give them some feedback about the data quality itself. In the app, for instance, right now, it's being done by a checkmark. If you take a picture of the sky and that has been taken into account by the data fusion algorithm, then you get the checkmark, let's say. And then exemplifying is, for instance, through storytelling. We are highlighting particular stories of users being involved in the project. And this is also something that we will do with the VRT here in Brussels. Then some last pictures to conclude. So on the website, you can also find different modules that you can download. If you would be interested yourself to organize a workshop around the sensors, there is a buying list available. There is tutorial material available and some feedback forms, so you can start independently organizing workshops. Actually, you don't need our help in this way. To show you the tutorial material, this was a workshop already organized in Norway with some students. And also, some videos have been recorded to help you assembling the sensors on your own. So that's it. If you would have any questions. Thank you for the presentation. My name is Damian Jacques. I have a question maybe for all the speakers. But when I saw your presentation, it made me think about it. It's about the spatial bias of variability of air pollution. So when you are measuring using home sensors, I guess that within the street you have a lot of variability. So my question is, yeah, maybe you have spatial bias when you are looking to two different streets when you will have a home sensor that will be at the top floor, for example, and one on the ground floor or in the middle of the street. So it wouldn't be more interesting to use only mobile sensors to have very accurate air pollution measurement. Yeah. Well, the first thing I maybe have to say then is that the exposure to air pollution is very context-specific, time-specific. So if you're about to install a sensor in the backyard, in the front yard, it might yield different results. Right now, we see a more interest in having the the Hacker home sensors as it is easier to install than the Hacker mobile sensors. So right there, I can't answer your question right now. If it wouldn't be better, if you would all have Hacker mobile sensors. Yeah.