As a fully digital energy service provider, enspired uses data at the heart of all its trading operations. To improve profitability of our clients’ power assets, we deploy self-learning trading strategies in intraday markets. Our innovative trading platform caters for the seamless integration of latest AI-technology with real market data and live asset-trading, to augment the capabilities of human traders and thereby reach a new level. We build on a unique combination of technology and energy-market know-how, which will take the performance of flexible power assets to a new level.
The advancing expansion of renewables is generating more and more volatility on the short-term electricity market and the need for a correspondingly high degree of flexibility as a counterweight to keep the grid stable and enable the energy transition.
The entry barriers for medium and smaller market participants to provide flexibility from their existing power plant portfolio remain too high due to market access costs, the necessary technological capabilities and the required 24x7 shift operation.
Even some of the large, established energy companies are unable to process the relevant data in real-time in order to use and market flexibility in a targeted manner. The interdependencies in European power grids and our liberalized markets create a level of complexity that is very difficult to master with traditional methods of analysis, optimization and manual dispatching of power assets, which it still the de-facto status quo across wide parts of the energy landscape.
To operationalise our concept, we engage in intensive research & development in several areas:
- artificial intelligence in short-term electricity trading for trading and optimisation decisions is not yet in productive use in practice and has been little researched;
- some of the proposed solutions, such as the combination of Reinforcement Learning (RL), deep neural networks (DNN) and other machine learning (ML) algorithms that are more suited for dealing with noisy data (e.g. random forests), are generally not yet well advanced;
- regulatory framework conditions in energy trading require the traceability of trading decisions in a level of detail that is not yet available with current AI methods (comprehensible AI; "explainable/interpretable AI").
We have proven that the market entry barriers and optimisation challenges can be overcome with the use of Artificial Intelligence (AI) and a tailored architecture. Through complete automation and scaling via generalisable models, we gained a significant competitive advantage over traditional energy trading service providers. The resulting increased profitability of flexible assets benefits our customers and can be offered at low cost due to the high degree of automation.
Read more about this project