The recent interruption of the electricity supply in Spain has highlighted the urgent need to improve predictive capacity and strategic decision making in energy systems. In this context, the technology developed by the Spanish startup Aygloo is positioned as an essential ally for network operators and energy responsible.
Who is Aygloo?
Aygloo is a Spanish technology company specialized in explainable artificial intelligence (XAI), an emerging discipline that seeks to make the results of AI models transparent and understandable. Its platform is integrated into the existing systems of the client, without replacing them, and provides traceability, detailed explanations and advanced analysis tools that allow technical and business teams to understand and audit each prediction. The development of this Aygloo technology has the financial support of the Center for Technological and Industrial Development (CDTI).
The explainable to anticipate critical events
Aygloo’s claims are backed by numerous scientific studies, including a recent one that demonstrates how explainable artificial intelligence technology (XAI) improves the prediction of wind energy by offering a detailed understanding of prediction models, study developed by the Laboratory of Renewable Wind and Energy Engineering (EPFL_WIRE) of the Federal Polytechnic School of Lausana (EPFL)
Aygloo has developed technical innovations such as dynamic subrogate models and intelligent segmentation algorithms
The use of this technology is highly relevant to reduce the energy dependence of Spain and the EU, optimizing the integration of renewable sources into the system.
The Aygloo platform offers unique functionalities:
- Detection of critical segments where the model fails or generates uncertainty.
- Simplified twin models that simulate alternative scenarios, including exogenous factors, without resentment
- Multiscala explainability, which offers analysis from a global vision to the schedule detail.
- Impartiality and robustness analysis, including the detection of biases and imbalances in the data.
- Simulation of “What-F” and counterfactual scenarios, to analyze how variations of factors (such as temperature or price of electricity) affect predictions.
- Complete traceability of decisions, guaranteeing that each prediction is understandable and auditable.
Aygloo has developed technical innovations such as dynamic subrogate models and intelligent segmentation algorithms capable of extracting powerful insights that can prevent facts as extreme and the recently suffered in the Iberian Peninsula.
Transform data into a decision
Unlike approaches that only prioritize precision by “black box” type models, Aygloo provides the intelligence layer that transforms predictions into useful, explainable and actionable knowledge. Its proposal allows operators to anticipate problems, adjust their existing models and make informed decisions even in conditions of high uncertainty, such as those lived during the recent blackout.
As the CEO of Aygloo, Ignacio Gutiérrez Peña underlines:
«Our technology transforms predictive models into strategic tools, allowing those responsible for the electrical system to anticipate critical events such as the recent blackout. It is not just about predicting, but to understand why, how and when something will happen ».
In short, Aygloo does not compete with the traditional models of Machine Learningbut the power. Faced with an increasingly complex and volatile energy environment, solutions such as yours allow us to move from an isolated prediction to a complete, auditable and actionable history that reinforces the resilience of the electrical system.
In addition to its approach in the energy sector, Aygloo leads other innovative R&D projects, such as one with Europa Press and the Autonomous University of Madrid for the development of a tool that detects false news and explains why they are, applying techniques of the explainable.
