Project Overview

Collaborative project with the NYPA to identify interruptions in an underwater cable between Long Island and mainland New York. Predict evolving faults and disturbances with as long a time horizon as possible.

Process

Applying AI/ML techniques to historical data from various sources. Identifying small anomalies, deviations, and patterns to predict evolving faults and disturbances as far in advance as possible.

Outcome

Initial cable faults on buried & underwater cables could be predicted

• 2 months in advance with an 80% level of confidence
• 24 hours with a near 99% level of confidence.

Eneryield IntelliView™ goes beyond the capabilities of conventional power grid analysis techniques and identified data correlations, indicators, and characteristics. The AI/ML techniques tailored to their needs, enable NYPA to achieve high confidence in predicting faults and identifying key operational drivers.

Project Overview

Together with ABB, we want to make our contribution to an increasingly electrified society with a higher reliance on an uninterrupted power supply. The goal is to: Reduce the number of interruptions in the power supply, enhance fault localization, isolation, and supply restoration processes. Minimize the duration of faults when they do occur.

Process

• The solution is based on an explainable AI, XAI, machine learning-based fault prediction designed to foresee incipient faults and identify how the conclusions are drawn, avoiding false positives.
• Proactively ensuring power supply reliability, enhancing prediction and prevention of faults, ultimately improving customer experience by minimizing potential outages.
• Utilizing existing measurements of current and voltage, eliminating the need for additional sensor installations.