Eduard Saller presented on forecasting demand for a car sharing fleet. The presentation discussed (1) using machine learning models like random forest regressors on historical vehicle pickup and drop-off data along with other data sources to capture trends and forecast demand, (2) monitoring fleet coverage gaps in real-time, and (3) using the forecasts to optimally distribute vehicles to improve availability and reduce idle time. Future work could include incorporating more context data and expanding the approach to other applications like electric vehicle charging infrastructure planning.