Integrating AI Models for Voltage and Current Monitoring in Autonomous Mobile Robots to Prevent Power System Blackouts

Autonomous Mobile Robots (AMRs) are increasingly deployed in industries for tasks involving navigation, transportation, and automation. However, fluctuations in voltage and current can cause significant performance degradation or hardware damage in these systems. This project proposes an AI-based predictive monitoring system designed to detect voltage and current anomalies in AMRs using machine learning and deep learning models. The solution integrates a Flask-based web interface for real-time prediction, model evaluation, and anomaly detection using Z-score thresholds.

A combination of regression algorithms—including Random Forest, XGBoost, Decision Tree, Gradient Boosting, and K-Nearest Neighbors—was evaluated based on metrics like Mean Squared Error (MSE) and R² Score. The Random Forest model, showing superior accuracy (MSE: 0.0027, R²: 0.9985), was used for final predictions. User inputs such as reactive power, voltage, intensity, and sub-metering values are analyzed to detect potential abnormalities. Predictions are contextualized using historical values and Z-score computation to classify output as “Normal” or “Anomaly.”

The system supports secure login, registration, dataset upload, and selection of different algorithms for evaluation. This intelligent, data-driven approach enables proactive fault detection, preventing electrical failures and ensuring operational efficiency in AMRs.

Keywords:

Autonomous Mobile Robots, Voltage Monitoring, Current Monitoring, Anomaly Detection, Machine Learning, Random Forest, Flask, Z-score, Predictive Maintenance, Energy Efficiency.