SOME takes centre stage at the AI Congress 2025 in Barcelona

On 22 and 23 October 2025, the eleventh edition of the AI Congress was held in Barcelona. The event, organised by CIDAI and Eurecat as part of the Catalonia AI Strategy promoted by the Government, brought together professionals and organisations from the fields of AI and data. The congress offered the chance to discover new AI solutions applied to business, as well as to explore trends, best practices from pioneering companies, new developments, and inspiring success stories.

On the second day, SOME presented its project in one of the congress sessions. The session entitled “From Proof to Deployment: Scaling AI Successfully”, held in a packed room, showcased how various companies have introduced AI solutions into their production environments with real business impact. In this session, Oriol Canillas and Ernest Ceballos from SOME shared their experience at SOME developing a solution based on models such as LSTM and HMM. Other members of SOME’s IT team also attended the congress as audience members.

The project arose from the need to improve production efficiency and reduce the impact of defective parts (NOK), which often create bottlenecks and delays in production. Furthermore, in some cases, the exact causes of these defects were unknown, making it difficult to apply corrective measures. Faced with this challenge, the SOME team initiated the development of an artificial intelligence solution capable of predicting when a defective part might be produced. To achieve this, large volumes of data generated during production were collected and analysed. A significant part of the project time was devoted to preparing and cleaning this data in order to ensure the quality required to train a reliable predictive model.

The solution is currently in a pilot phase within a real production environment and is expected to be fully deployed in the coming months, in a sector as demanding as the automotive industry.

With this initiative, SOME demonstrates the potential of artificial intelligence applied to industry to improve processes and outcomes.