Catalan startup Process Talks and Basque industrial group Danobatgroup have collaborated to develop a tool based on generative artificial intelligence (AI). This solution facilitates access to information generated by industrial machine sensors, allowing workers to query data in natural language and automatically identify the most suitable analyses.
The project aims to simplify the analysis of large data volumes, combining Danobatgroup's expertise in advanced manufacturing and industrial maintenance with Process Talks' experience in generative AI. The new tool enables users to ask questions through a conversational assistant, and the AI identifies the best way to visualize the information, automatically generating analysis graphs.
“"Machines have numerous sensors that constantly generate information. Until now, technicians worked with tools and dashboards, which required pre-configuring the graphs they wanted to consult."
For Danobatgroup, this project has validated the potential of generative AI applied to predictive maintenance. According to Iván Castro, Head of Innovation, they had been working on this technology for a long time and were able to develop a first pilot project with Process Talks, confirming they are on the right track.
One of the key advancements is that workers themselves can perform an initial data analysis, without constantly relying on analytics specialists. This tool allows them to make initial assessments without needing to be data analysis experts, detecting anomalies more quickly and streamlining decision-making.
This project also highlights BIND's role as a hub between technological startups and industrial companies. Process Talks emphasizes the opportunity to develop a solution tailored to real industry needs, and Danobatgroup views the BIND program as a strategic tool for incorporating innovation and exploring new technologies alongside specialized startups.
