Marcos Blanco Fernandes Email & Phone Number

Albertville, Alabama, US

• Spearheaded migration of machine learning technique for autonomous freight train driving strategy development, in 3 months since conception to field test
• Evaluate driving strategy performance and main controller logic prior to field deployment and production
• Introduced a pioneer methodology for generating synthetic data enhancing performance of AI models
• Reduced time to provide and insert human feedback into learning cycle from 8h to 1h for a single train through developing and implementing GUI tools
• Lead a team of Machine Learning, Data, and Software Engineers (~10)
• Deployed ML models in North America Class 1 Railroad capable of 25% reduction in fuel consumption and increase in train safety

Albertville, Alabama, US

• Trained and deployed Machine Learning models for decision-making process and control strategy for autonomous freight train driving
• Developed, implemented, and deployed custom Supervised, Unsupervised, and Reinforced Machine Learning training algorithms
• Designed main controller logic integrating core Machine Learning models to decision-making system
• Enhanced performance of vehicle dynamics simulators for freight trains up to 0.9 ms/s with transient forces calculation
• Calibrated physics engine parameters with real field data, improving accuracy by 15%
• Architected and implemented end-to-end digital pipeline that process unstructured data and validates driving strategy performance utilizing train dynamics simulation environment (digital twin)

• Created proprietary software for freight train driving simulations, providing railroads decision-making tools
• Implemented machine learning algorithm to predict maintenance needs and optimize reliability of freight train systems
• Designed and programmed a centralized database system to store and manage critical operational data for freight train operations
• Consulted in an analysis of operational simulations for freight trains

Curitiba, Paraná, BR

• Architected, programmed, and tested pilot software for assisting Locomotive Engineers in driving strategy to improve train velocity achieving 3% reduction in transit time (semi-autonomous)
• Analyzed human adoption and adherence to semi-autonomous systems in railroad operations
• Developed train dynamic digital simulators to estimate train speed, fuel consumption, and in-train forces, aiming improvements on operational performance
• Increased cargo volume by 40% in a freight train type, resulting in better energy efficiency, through studies on maximum rail and coupler load capacity and field tests on network
• Led multidisciplinary teams of Specialists, Analysts, Supervisors, and Locomotive Engineers in different departments

Curitiba, Paraná, BR

Master'S Degree, Mechanical Engineering

Executive Programme, Project And Infrastructure Finance

Engineer'S Degree, Mechanical Engineering

Aerospace Engineering