Welcome to EMAST

Welcome to our project Homepage. Here you can gather information about EMAST (Energy Market AI for Smart Meter Tariffs).

Short description

English
The Stadtwerke Klagenfurt currently only have static tariffs. There is data from around 320 smart meters (which transmit quarter-hour values) and from two energy markets, but this data is not yet used to create dynamic tariffs. The two energy markets are the futures market and the spot market, which determine the electricity price.

The quarter-hourly values from smart meters in Klagenfurt and energy market prices are analyzed using algorithms and artificial intelligence. This creates an improved tariff for the end user and the Stadtwerke Klagenfurt benefit from improved electricity purchases. The result is then saved in a database and is available as a file.

Our project is being processed and created at HTL Mössingerstraße

The business partner for this project are the Stadtwerke Klagenfurt


German
Die Stadtwerke Klagenfurt haben momentan nur statische Tarife. Es gibt Daten von ca. 320 Smart Meter (die alle Viertelstunden-Werte übermitteln) und von zwei Energiemärkten, jedoch werden diese Daten noch nicht zum Erstellen von dynamischen Tarifen verwendet. Bei den zwei Energiemärkten handelt es sich um einen Termin- und Spotmarkt, welche den Strompreis bestimmen.

Es werden die Viertelstunden-Werte von Smartmetern in Klagenfurt und Energiemarktpreise durch Algorithmen und Künstliche Intelligenz analysiert. Dadurch wird ein verbesserter Tarif für den Endnutzer erstellt und die Stadtwerke profitieren von einem verbesserten Stromeinkauf. Das Ergebnis ist dann in einer Datenbank gespeichert und ist als Datei verfügbar.

Details

Here is a short overview of how our projekt works and what components we use.


Flowdiagram

Flowdiagram of EMAST

The flow of EMAST can be broken down in 4 steps:
- Step 1: Downloading information from APIs and reading data from the STW database
- Step 2: Formatting the data from web and validate all data.
- Step 3: Store the input data, process it through the AI and store the output.
- Step 4: Visualize the output and input, make it accessible through an API and export it as a file.

trending_up Grafana

Visualization

Grafana is tool that is used to analyze and visualize data from a database. EMAST is using it to visualize the input data and also the output data. You can test the visualization by clicking at the Grafana logo down below.

AI Structure

Structure of the AIs

We need two AIs to accomplish our goal, because there are two energy markets the Stadtwerke Klagenfurt get energy from. Each AI predicts the amount of energy for one of those Markets. With the consumption and user data from the past few years as input data the AI has a good foundation, but to train an AI it needs to be given output data. In this case that would be how much futures and spot energy should be bought. This output data must be calculated by hand or with an algorithm.

Training and Re-Training

The AIs will have to train one time with the whole data set to get a foundation. After that it will be Re-Trained daily. That is because we get new smart meter values every day. The Re-Training has the purpose to make the AIs more accurate.

Interfaces

The export of the data is done through two different interfaces. There is a CSV file for importing data into the STW database that is updated every day. In addition, a Rest API interface is available where data can be retrieved through simple GET requests. The Documentation of the Rest API can be reached by clicking at the Postman logo.

Technologies we are using

Access our source code and get further documentation

Our Awesome Team

Wait, who said we are awsome?!

Alexander Jahrer

AI Development

Wenn du blind bist kannst du net schaun!
~Stefan Jöbstl

Stefan Jöbstl

Database and Backend

Debugging is being the detective in a crime movie
where you're also the murder.

~Filipe Fortes

Prof. Dipl.-Ing. Dr. Daniel Wischounig-Strucl

Mastermind of HTL Mössingerstraße

Get some probably banana.
~Dynamic Daniel