With the proposed challenge, we developed a pilot of a technological solution, an application – PowerBI (Scorecard and AI Modeling).
To solve the challenge, we created a pilot of an Artificial Intelligence solution capable of dividing and mapping railway sections that presented the probability of track deformations (buckling).
Fifteen rules were developed to identify the events that contribute to this type of problem in railways. Some of the main ones include temperature, rail type, ballast and fastener, date of last maintenance, season of last maintenance, among others.
Based on these rules and scores, we delivered a Power BI application capable of displaying the scorecard results, combined with an artificial intelligence model.