“Using historical and real-time data, IA4 students have created a system capable of analyzing and predicting traffic accidents throughout Spain.”
Jon Medina, Michael Jiménez, and Mikel Martínez, students specializing in Artificial Intelligence and Big Data at the Somorrostro Training Center, have started the academic year with an ambitious project as part of the so-called Zero Challenge. This first challenge consisted of working with any series of data, analyzing it, and exploring possible applications or business ideas based on it.
They decided to focus on historical and real-time data on traffic accidents throughout Spain, with a clear objective: to unify the various existing sources and create a tool that would enable them to visualize, analyze, and predict road accidents.
“We saw that the DGT had a map with incidents, but it excluded the Basque Country and Catalonia, because these are regional jurisdictions,” they explain. “So we looked for APIs from the three sources and worked to integrate them into a single dashboard.”
The result is an interactive dashboard that shows accidents recorded in Spain, the Basque Country, and Catalonia in real time. Thanks to tools such as Node-RED, InfluxDB, and Grafana, the students have managed to process and visualize thousands of records that allow them to see, for example, which roads have the highest concentration of accidents or which are the most active black spots at any given time.
“In Spain, there is approximately one accident per minute. In the Basque Country and Catalonia, there is one every half hour,” they say. “With all this information, we can analyze patterns and even predict risk areas based on weather, holidays, or holiday traffic.”
In addition to its social utility, the group has identified a potential business value for the project. “Predictive services could be developed for infrastructure companies, insurance companies, or even mobile applications that warn of dangerous areas,” they point out. “And all without the need for large servers or investments, because the data is managed in the cloud.”
Although the challenge is over, the three agree that this first project has been an excellent introduction to specialization.
“It’s very interesting because of how practical it is,” says Jon. “We’ve gone through the whole process: choosing the data, processing it, visualizing it, and finding its usefulness. It’s just what we would do in a real company.”
