The collaboration between Unusuals and Redeia dates back to 2020, when the startup began working with Red Eléctrica’s line Maintenance team to develop detection models for the (DALIA) anomaly detection project—an innovative solution for creating and managing digital models for linear infrastructure.
This relationship has been further strengthened by Elewit’s corporate venture capital investment in the partner company. Through this investment vehicle, Elewit seeks technology-based startups driving efficiency, technology monitoring, new business development, and other strategic priorities aligned with the group’s key challenges. In this regard, Unusuals' solution addresses the need for digitising grid asset maintenance and integrating AI into inspection and maintenance processes.
Founded in 2016, Unusuals set out to automate infrastructure maintenance. To that end, the company supports visual inspection processes in the maintenance of critical infrastructure, leveraging AI models to detect potential defects and anomalies, thus significantly accelerating the inspection process.
Additionally, Unusuals’ technology incorporates a dedicated software layer for critical infrastructure and uses synthetic data to train AI models. This approach enables the automation of existing asset maintenance processes, reducing costs and minimising risks.
Elewit’s investment in Unusuals is driven by:
- Scalability and growth potential of the technology.
- High-quality detection algorithms.
- The expertise and talent of the team.
- A unique solution for synthetic image generation.
While Unusuals previously collaborated with Elewit—participating in its first venture client programme and winning the Elewit Data Challenge in 2020—the startup is now actively involved in the DALIA project. In this project, it provides AI models for automated detection of visual anomalies in images captured during overhead line inspections, contributing to digital asset inspections.
DALIA aims to develop an AI-powered tool that assists professionals in analysing inspection images. This tool can even be trained using real images or synthetic data that simulate anomalies found in overhead power lines
Unusuals plays a key role in this process, offering expertise in synthetic image generation for AI training and working closely with Red Eléctrica’s team to validate photorealistic images of environments and defects.
As a result, Redeia’s partnership has enabled Unusuals to validate its technology, increase automation levels in visual inspection processes, and expand into new markets and sectors beyond its initial scope.