View the final Project Report of #class21 Team "ElecTUM"
Our research question
How can we make the TUM more sustainable after the COVID-19 pandemic?
Develop a model of student scenario-based behaviors that influence our energy consumption with the aim of encouraging the university and its students to take climate action based on these findings.
What have we been working on?
Team elecTUM was a research group composed by 8 students from 7 different courses and 4 nationalities! Having started as “Team Climate”, our goal always was to investigate a disruptive situation that had an impact on the climate. After several weeks of discussion rounds, we decided to focus our project on the effect of different types of lecture on the energy consumption of TUM students. In this context, we were specifically interested in finding out what hybrid format between online and on-site lectures maximizes the energy efficiency of university teaching. To find out the answer to this question, we implemented a calculator that predicts the energy consumption of TUM lectures, depending on the way these are taken. Since our calculator is based on real/realistic data, we divided our team into three subgroups, from which two are thought to gather the necessary data for the calculator.
Team data handling
The first subgroup (“data gathering/analysis”) was responsible for collecting the fundamental data for the model calculation. Regarding this aim, this subgroup investigated the electricity consumption of a model lecture hall and a model student apartment, the transportation of students, and the streaming of online lectures. They worked on gathering the data for the model lecture hall, the model student apartment, and streaming. For the electricity consumption of lecture halls, the subgroup decided to use one of the Interims lecture halls as a representative for a modern lecture hall. As the TUM does not monitor the electricity consumption of the lecture halls in detail, the group manually calculated the electricity consumption of an Interims lecture hall. Regarding the model student apartment, the group used the Studentenwerk apartments as basis. Since the energy consumption data collected from these apartments were also not detailed enough, they manually calculated the electricity consumption based on the student’s streaming device (to be investigated by survey), internet router, and light infrastructure. For the latter two, the group was in contact with the dorms to identify the technical details. Lastly, for identifying the energy consumption of the video streaming, the subgroup decided to do an extensive literature research. They defined the different parts that require energy during the streaming process to decide what should be included in our calculation and what should be asked to the students by the survey group.
To complement the first subgroup, we created the “survey” group for the purpose of student-based data collection, specifically on student transportation, device use, energy consumption at home etc. The subgroup outlined a plan that included scheduling, preliminary survey design, internal and external review rounds, small group testing, and publication. They decided on a platform on which to design the survey (evasys.de) and were given access. Since it was important for data collection to have a representative number of participants, this subgroup worked on a plan to reach a substantial range of students and staff. After the first draft, they sent the survey to our other group members and our mentors for review while updating the design. Their goal was to make the survey as simple and engaging as possible to encourage more participants to complete it while still providing the necessary data for our calculator. After the first draft was completed, this subgroup used a small group of participants (each group member sent the survey to around five people she/he knows) to test the survey and provided us with feedback.
Lastly, the subgroup “calculator” was created. To analyze the gathered data, a calculator in form of a web-application was being developed, which was made publicly available towards the end of the project. It calculates the overall energy consumption of both online and on-site lectures through a linear model developed specifically for this project, as well as a mix of frequentist inference and linear error propagation to estimate the uncertainties of our results. This model was made accessible through an interactive web interface, allowing for flexible user input of variable parameters.
Maryam Tatari | Sebastian Zäpfel