Azul Brazilian Airlines is planning to resume its artificial intelligence-driven maintenance projects following its exit from Chapter 11 bankruptcy protection. The airline, which made significant advances in using machine learning for predictive maintenance before filing for bankruptcy in May 2025, aims to leverage this technology again to reduce aircraft downtime and improve operational efficiency.
Prior to restructuring, Azul utilized AI to predict and prevent aircraft on ground (AOG) events, reducing component failures and increasing fleet availability. The airline's director of planning, André Américo, reported that their AI system was able to forecast and mitigate about 150 events per month, resulting in weekly savings of approximately $6 million. The project involved digitizing maintenance logs, translating technical codes, and using language models to analyze aircraft condition data.
After raising $850 million in new investments, Azul exited bankruptcy in February 2026, with plans to advance this AI work. Américo emphasized that the collaboration with external AI partners, such as AE Studio, helped break down departmental silos and foster a more integrated approach. The airline expects to use data more effectively in scheduling maintenance, managing revenues, and optimizing fleet deployment, aiming for a better balance between scheduled upkeep and unexpected AOG events.
Azul's AI initiatives also contributed to better engine management, especially during supply chain disruptions. Américo shared that by analyzing engine performance data, the airline could shuffle engine serial numbers among aircraft to maximize operational uptime. The airline’s goal now is to expand these predictive capabilities further, incorporating more data sources and refining their models to enhance reliability and reduce maintenance-related delays.
