TerraCast

challenge

Challenge 4

brief

TerraCast developed TerraCast, a machine‑learning based forecasting approach that combines data quality checks, classification, and regression models to predict schedule delay risk and likely lateness across energy projects, supported by dashboard‑ready outputs.

Please be aware that this content was generated follwing an automated review so may not be perfectly accurate; refer to the original challenge brief and team files for authoritative information

key outcomes

Expected to help project teams identify delay risk earlier, quantify likely lateness in days, and improve confidence in schedule forecasts through clearer data quality and feature‑trend insight.

important files

• TerraCast_Grounded Planners_User_Guide.docx: Explains the end‑to‑end TerraCast approach, including data quality checks, classification of on‑time vs late projects, and regression‑based delay prediction.
• PowerBI- raw data/in_progress_predictions.csv: Prediction outputs used to visualise forecast lateness and risk status in dashboards.
• Model - Raw Data/Completed_Energy_Projects_Lifecycle_Dataset.csv: Historical dataset underpinning model training and feature‑trend analysis.

details

team: TerraCast members: tbc topics: solution-centre, hack25, challenge4, python, scikit-learn, logistic-regression, linear-regression, power-bi, schedule-forecasting, machine-learning, predictive-analytics, data-quality, project-controls, delivery-performance technologies: Python, Scikit-learn, Logistic Regression, Linear Regression, Power BI