A lot of activity is being devoted to studying issues related to energy consumption and e ciency in our buildings, and especially on public buildings. In this context, the educational public buildings should be an important part of the equation. At the same time, there is an evident need for open datasets, which should be publicly available for researchers to use. We have implemented a real-world multi-site Internet of Things (IoT) deployment, comprising 25 school buildings across Europe, primarily designed as a foundation for enabling IoT-based energy awareness and sustainability lectures and promoting data-driven energy-saving behaviors. In this work, we present some of the basic aspects to producing datasets from this deployment and discuss its potential uses. We also provide a brief discussion on data derived from a preliminary analysis of thermal comfort-related data produced from this infrastructure.
in several countries. Recently, there has also been an increasing interest in the design of energy awareness educational activities centered around IoT-enabled lab approaches. In this work, we argue that producing datasets from public buildings, such as school buildings, has several bene ts and such datasets can be used in a number of useful ways.
As our example for realizing such an approach, we base our work on the Green Awareness In Action (GAIA),
project. GAIA [
The European Union has been collecting data for several years now from projects related to energy consumption
and e ciency and making them public through the BuildUp portal [
Regarding uses of datasets in an educational setting, there is a growing interest in utilizing such data in recent
years, especially in the context of makerspaces, hackerspaces and other similar groups. There are some examples
where IoT-driven educational activities have been performed with the additional objective of increasing students'
awareness of societal challenges. [
Thermal comfort is one of the most important parameters for adapting the conditions inside a public building to make it \better" for its everyday users. Especially in school building, where users are mainly children and teenagers, thermal conditions are of utmost importance as they do not only o er comfort to users but a ect the performance of students and the educational process itself. Thermal comfort is de ned as the condition of mind that expresses satisfaction with the thermal environment and is assessed mainly by subjective evaluation, but can be objecti ed at some extent. The main factors that in uence thermal comfort are those that determine heat gain and loss, namely metabolic rate, clothing insulation, air temperature, radiant temperature, air speed and relative humidity. Psychological parameters, such as individual expectations, also have some e ect but cannot and will not be measured in our context.
We focus on a single metric, the PMV model, that was developed by P.O. Fanger [
Based on the data collected from speci c school buildings we monitor in GAIA, we present here the calculated PMV value, as well as temperature and relative humidity data from the same building for a typical day. For the rest of the parameters that are not directly monitored we selected and used xed values (i.e., air speed is 0 since we refer to indoor measurements). The data from a single building are presented in Fig. 2. From the gure, it can be seen that for this speci c building PMV is within acceptable limits; going over similar data over greater time periods can provide more useful insights with respect to how conditions inside such a building play out.
Similar data can be extracted from other buildings of our network and the conditions in each of them can be easily evaluated even in real-time. A comparative view of the PMV in two buildings can be seen in Fig. 3 where the PMV value for the morning, working and evening hours is presented for the aforementioned school and another one located in a colder region. Generating such data and exporting them as datasets that can be shared with other researchers can help to identify other approaches, or ideas that can further our understanding and increase the practical usefulness of IoT installations. They can also be used by e.g., local government agencies to detect di erences between the way buildings in the same region behave and spot the \weaker" ones, that show greater problems, or spot the best ones that should serve as an example of operation for the rest.
Moving on to more general comparisons between school buildings and detecting overall issues in the way school buildings operate, in Fig. 4 we show the percentage of time during which temperature was outside comfortable levels (i.e., above 25 C or below 19 C) during wintertime. This chart represents the conditions inside 21 school buildings in Europe, and it is interesting that there are big di erences in these results. Apart from 3 schools that appear to have essentially 0% time outside of comfortable levels, many schools appear to have serious issues with the handling of indoor temperature during this period of the year. 4
We now discuss related educational activities that have been proposed targeting schools and integrating with the existing computer science curricula, and that utilize datasets from educational buildings. The pedagogical goals aiming at increasing students' awareness on energy topics, in GAIA, are: awareness, observation, experimentation and action. These goals can be used to inform IoT-enabled education activities, which can focus on helping students and teachers to get a better understanding of the energy-related processes taking place inside their school building, how their actions and routine contribute to their power consumption, or how they compare with other similar schools in the same or other countries.
In Fig. 1, several instances of IoT hardware used in actual hands-on educational activities are shown. Such devices are used inside the school to generate the data that can be then used by students and teachers to provide a better understanding of their natural environment surrounding. They also give students an understanding of the electrical and electronics that are associated with sensing and computers. E.g., they report almost in real time on temperature and humidity inside the class. A key aspect here, and in related datasets, is granularity and responsiveness; in some cases, available datasets focus on datasets with e.g., a granularity of 30 minutes. This does not mix very well with the available time resources in schools, where students would like to see the immediate e ect of their actions in some cases.
Knowledge and experiences gained from these courses, helps students to feel more comfortable with technology and the so called maker movement and build their own projects, either physical, using paper and paint, or more technical, like robotics and DIY both as part of their school curriculum or in their free time. An example of such an outcome can be seen in Fig. 5. It is an interactive installation that uses IoT data from school buildings, like overall power consumption, temperature or noise levels from the classrooms and visualizes it both on the paper-built control panel and on the connected monitor using a simple website. This interactive installation is easy to build and can be fully customized based on the respective school building and available sensors.
Having tools and data available, such as the ones mentioned above, opens up many possibilities for educators. Schools and educators can utilize them in the context of in-class activities according to particularities of their own school environment. Currently, there are few \o cially-approved" educational activities focused on sustainability. However, there is always room for inserting such aspects into the context of existing curriculum activities. Furthermore, there is a growing awareness among young people with respect to climate change and the need to act to mitigate its e ects, and school environments are a very good candidate for sustainability awareness activities. In the case of the installation depicted in Fig. 5, we have seen that it can be a starting point to make students aware of aspects that did not interest them in practice before, or were completely vague to them.
In this sense, the existence of school building data sets can be utilized in the context of statistics, physics, or
natural science classes. Such data series provide the opportunity to study natural phenomena, e.g., the e ect
of building orientation or materials' use on school building energy consumption behavior, or examine the e ect
of the presence of students and teachers across di erent parts of their school and study energy consumption
patterns. In practice, we have seen from our experience in the GAIA project educators using such possibilities in
various ways; even educators that do not have a technical background can nd ways to integrate them into their
work ow. The openness and availability of the data in the GAIA provided this kind of exibility. Essentially,
educators can use the available data as the foundation upon which they can work together with their students,
serving two aspects at the same time: i) working with data that relate to the students' everyday surroundings,
and ii) working in a sustainability-related context. Both have the potential to make in-class activities more
engaging and motivate students to become more activated in environmental issues. In [
Overall, in recent years there has been a lot of interest in terms of producing and utilizing datasets from buildings. This is more or less expected, since you cannot control what you cannot measure, and moreover it is not easy to improve or optimize something for which you do not have a clear picture. In this context, it is a fortunate event that recent advancements in IoT, data analytics, machine learning and cloud computing can help us in understanding processes related to the operation of public buildings, especially given climate change and the need to become a more sustainable society as soon as possible.
This issue nds a fertile ground in the case of educational buildings: datasets that are produced using e.g., an IoT substrate can nd many more uses than those of a typical public, non-residential building. Based on the examples brie y discussed in this paper, we can argument that the existence of open datasets from IoT-enabled school buildings provide plenty of opportunities to work on multiple \fronts". E.g., they can immediately be utilized in an educational context, providing students with something that relates to their immediate environment, and which can be used as the input in educational activities. We have discussed brie y in this work some existing options and their potential uses. On top of that, they can also be used as input for more generic uses, e.g., understanding energy consumption patterns, testing algorithms for energy disaggregation, etc. We believe that there is a lot of area of potential growth in this domain, and we will be working in the near future to provide more substantial and analytic datasets to the community for such purposes.
This work has been partially supported by the \Green Awareness In Action" (GAIA) project, funded by the European Commission and the EASME under H2020 and contract number 696029, and the EU research project \European Extreme Performing Big Data Stacks" (E2Data), funded by the European Commission under H2020 and contract number 780245. This document re ects only the authors views and the EC and EASME are not responsible for any use that may be made of the information it contains.