Term 252

Forecasting ESP Failures Using AI & ML Models

Project Type: Self-Initiated

Project Description

Electrical Submersible Pump (ESP) failures create planning uncertainty for annual budgeting, spare-parts procurement, and maintenance readiness. This project develops a fleet-level AI/ML forecasting system that uses historical ESP installation, failure, production, and validated electrical records to estimate expected ESP failures over the next 12 months. The system combines data validation, lifecycle-based modeling, fleet-level aggregation, and a user-facing dashboard/reporting workflow for PETE and EE planning users. The final output provides a monthly breakdown and annual total of expected ESP failures to support budgeting, procurement timing, maintenance scheduling, and spare-parts readiness.