In this year's industry challenge by hessian.AI, techies and AI experts are invited to detect anomalies and thus contribute to the stability of the European power grid.
@hessian.AI: What is an industry challenge?
Great question! An industry challenge is like a hackathon, in which AI developers participate to solve a specific case using AI methods in a short period. It’s a real-world laboratory in which problem and solution providers come together. For this purpose, our application partner TransnetBW provides a previously unpublished data set. Participants have 7 weeks to try their hand at it. The challenge comes as a mix of work phases on one’s computer and events that create meeting points. At the end, the winner (teams) will be crowned. They can look forward to lots of visibility, applause, and a surprise. But eventually, everyone is a winner, as the solutions help to make the European power grid more stable and reduce greenhouse gas emissions.
@hessian.AI: What is the topic of this year's industry challenge?
TransnetBW has provided a super-relevant case. Because: The European interconnected grid is the biggest machine in the world. For it to work, power generation and consumption must always be in balance. This requires constant adjustment. TransnetBW operates the PICASSO platform to optimize the use of control reserves across Europe. In its main control center in Wendlingen, the company collects data from different countries and determines the optimal energy exchange. Machine learning methods can be used to detect irregularities that might harm this vast machine. This is where the industry challenge comes into play.
@hessian.AI: And what problem are the participants supposed to solve exactly?
To ensure that the PICASSO platform functions properly, information on imbalances, reserves, transmission capacities, and many other time series are continuously monitored. Sometimes these variables behave strangely, indicating to the system engineers at TransnetBW that they may need to take action. (AI) algorithms that recognize such anomalies quickly and precisely could help to identify and allocate issues in reserve use or energy exchange in the future.
@hessian.AI: Who takes part in a challenge like this and why would you do so?
First, the challenge is open to everyone. Therefore, participants are quite diverse. There are, for example, founders from the energy or AI service sector. By attending, they draw attention to their offering, measurably demonstrate their skills, and get in touch with potential customers.
Researchers also take up the challenge. For them, the main attraction lies in working with new and real data. Science rarely has access to such data sets because companies seldom publish them. The
TransnetBW data set is a new playground that will be available to scientists even beyond the duration of the challenge.
Finally, students who are keen on getting to know a practical case alongside all the theory also join in. And by the way, it also doesn't matter where you are based geographically: anyone with an internet connection can attend because the challenge is hosted via the Kaggle platform and is accessible worldwide.
@hessian.AI: Ok, I am hooked, what's next?
You're welcome, this way, please:
- Further information and registration here
- Online kickoff event on Dec. 14th, 2023 from 3 to 5 pm
- 7 weeks for doing magic
- Final event on-site @TransnetBW in Wendlingen (near Stuttgart) on February 14th, 2024
The challenge partners TransnetBW, Fraunhofer IEE, House of Energy, House of Digital Transformation and hessian.AI are looking forward to getting to know you and are excited to see what you can get out of our data!
