MunichHome

Self Made Heating System
The complete Tech Workflow
The Software we used

Inspiration

The recent energy shortage due to the Ukraine war forces us to rethink our inefficient way of consuming energy. Buildings are responsible for one-third of total energy consumption, and 71% of the German heating system operates inefficiently. Our solution is to provide open-source available software and cost-friendly hardware to optimize the heating process in buildings. This way, we guarantee an easy setup to contribute to climate protection and save energy costs without paying too much money on external services or spending too much time on building it alone.

What it does

The end goal is to make energy savings and room occupancy visible on a Dashboard. Therefore, we use ambient temperature and occupancy calculated by two ultrasound sensors. They check in which direction people go and increase or decrease the counted number accordingly. Data is transmitted via ESP32 Wi-Fi modules to a Raspberry Pi. It then uses the incoming data to manipulate the controls of the heaters through another ESP32. Furthermore, the sensor data is pushed via a server agent (Telegraf) to the Influxdb database. That's where the Grafana Dashboard fetches the data from. Additionally, we build a simplified web view to present the real time state of the system in a more end user friendly fashion.

How we built it

We use ambient temperature and occupancy calculated by an ultrasound sensor. Data is transmitted via ESP32 Wifi modules. Additionally, we offer a Grafana Dashboard to track energy consumption and the resulting savings.

Challenges we ran into

We used docker for smooth deployment process and to make the app portable. Also, all hardware devices are working in the same network. We implemented a complete end-user prototype. Running two ultrasound sensors in parallel brought unexpected problems such as signal interference and the need to use multiple trigger signals for getting the distance. Moreover, we had to deal with limited hardware for connecting the sensors, which resulted in us making more efficient use of the available ports. Making the occupancy algorithm work.

Accomplishments that we're proud of

Counting people in a room is complex and prone to errors due to the inaccuracy of ultrasound sensors. We successfully built a system for counting people in a room based on a state machine-based algorithm that we came up with at 3:30 AM (finally!).

What we learned

How to integrate sensors with IoT functionality and collaborate with team members from various backgrounds. Additionally, we learned to use rapid prototyping to prove a concept of a self-made automated heating system (rooms made out of paper).