Machine learning models can significantly enhance project planning and control by using both static and dynamic project indicators. Static indicators, such as network data (e.g., activity dependencies and durations) and resource data (e.g., availability and allocation), provide a foundational overview of the project. Dynamic indicators, which incorporate real-time progress data (e.g., activity completion rates, delays, and resource usage), allow the model to adapt and learn from ongoing project developments. These models can generate predictive schedules that simulate project progress under uncertainty, forecasting the final time and cost more accurately.

The accuracy of these predictions will be compared with traditional Earned Value Management (EVM) metrics, which are commonly used for project performance evaluation. By using both artificial and empirical project data in computational experiments, the effectiveness of machine learning models can be assessed against established methods, potentially offering more precise and adaptive project management tools for handling uncertainties.