Advanced Data Analytics for a Renowned WEC racing Team
Developed frameworks and pipelines to provide United Autosports with race strategy and driver performance assessments, including flags, passing, traffic management, consistency, and pace.
These advanced analytics were utilized by the client to strategically plan and prepare for high-profile endurance races such as the Le Mans 24 Hours.
Predicting Mission-critical Failures in Locomotives for a World-leading Rail OEM
Developed, deployed and monitored ten ML models to predict failures in freight locomotives from sensor readings.
These models helped four main class-I railroads prevent dozens of mission-critical problems and hundreds of downtime hours.
Forecasting of Ambulance Demand for the Largest EMS Provider in the US
Developed ML models to predict the number of ambulance transports within hundreds of geographic business units, aiming to improve fleet utilization and on-time performance.
This models achieved, on average, a performance 20% to 40% superior to existing statistical forecasting models.
Forecasting of Air Pollution Concentrations for the City of Queretaro (Mexico)
Led a full-stack team of data scientists and engineers to build an application for monitoring and predicting future air pollution levels in the State of Querétaro, Mexico.
This forecasting tool is currently in use, aiming to provide authorities with enough lead time to implement measures that reduce public exposure to unhealthy air quality levels.
Detecting Mining Equipment Problems for the World's Largest Copper Producer
Developed and deployed eight anomaly detection ML models for Chilean mining giant Codelco, to flag potential issues in key mining equipment (such as conveyor belts, coolers, and ESPs) based on sensor readings.
This project was estimated to save $28 million in maintenance costs and reduced unplanned downtime.