ENERGY SYSTEMS · ML · OPTIMISATION

42.24°N −8.72°W

Juan Boullosa

Machine learning & optimisation for energy systems — batteries, wind, and power markets.

VIEW WORK CV

ABOUT

I’m Juan, an engineer from Vigo, Spain. I work on software for energy systems forecasting, dispatch optimisation, and real-time control for batteries and wind farms. I did wind-energy research at ForWind and NREL, and co-founded Zentus, an energy-analytics startup, through Stanford’s sustainability accelerator.

VIGO OLDENBURG NINGBO BOULDER MORE ABOUT ME →

SELECTED WORK

Projects

VIEW ALL WORK →

FORECASTING / DEEP LEARNING

Spatiotemporal Wind Forecasting with Uncertainty

Transformer models predicting full probability distributions, so wind-farm control can act on uncertainty.

PyTorch TACTiS-2 FLORIS
CASE STUDY →

OPTIMISATION / MARKETS

Multi-Market Battery Dispatch with MPC

A rolling-horizon MPC engine and dashboard dispatching battery storage across ERCOT’s day-ahead and real-time markets.

Python MPC Plotly
CASE STUDY →

REAL-TIME CONTROL

Model Predictive Control in 550 Microseconds

An embedded quadratic-optimisation solver running MPC on cRIO/FPGA hardware, validated in wind-tunnel experiments.

C LabVIEW FPGA
CASE STUDY →

CONDITION MONITORING

Blade Lifetime Extension for Offshore Wind

Physics-informed blade condition monitoring — damage simulation, anomaly detection, and remaining-useful-life estimation.

OpenFAST ROSCO RUL
CASE STUDY →