United States – Avangrid, a subsidiary of Iberdrola, is harnessing the potential of artificial intelligence (AI) and machine learning to enhance grid management. The formation of a specialized “Data Science and Analytics” team composed of experts from diverse fields including astrophysics, healthcare, and finance underscores Avangrid’s commitment to innovating grid operations.
The seven-member team is tasked with developing three distinct AI and machine learning-based systems: Predictive Health Analytics, GeoMesh, and HealthAI. Each of these systems leverages existing data from Avangrid’s electricity grids to predict grid performance, assess equipment condition, and target areas requiring inspections and investment. The ultimate goal is to elevate the reliability of services for the 2.31 million customers across Avangrid’s subsidiaries, including Central Maine Power, New York State Electric & Gas, Rochester Gas and Electric, and United Illuminating.
Pedro Azagra, CEO of Avangrid, highlighted the utility’s transformative approach: “We’re reimagining what’s possible for a utility when it comes to data science and analytics.” The initiative aims to empower Avangrid to internally create machine learning models that enhance grid operations and customer experience.
The “Predictive Health Analytics” system takes a proactive stance by assessing the condition of substation equipment, such as circuit breakers, to preemptively replace components before outages occur. Unlike traditional replacement methods based on age or malfunctions, this system leverages numerous factors, including age, usage frequency, manufacturer, and maintenance notifications, to predict equipment health and longevity.
“GeoMesh” is a comprehensive mapping project that evaluates the strengths and vulnerabilities of Avangrid’s service areas, enabling performance forecasts for various scenarios. By segmenting service areas and analyzing data points such as weather conditions, outage history, vegetation density, and more, Avangrid aims to optimize its response during both regular and storm conditions.
The “HealthAI” initiative involves analyzing millions of high-resolution photos of the distribution system to catalog assets and assess their condition. This system’s focus lies in identifying visible assets and gauging their health, ultimately aiding in raising awareness of equipment status and identifying areas requiring attention. The AI system’s capacity to correctly identify grid equipment in images and subsequently assess their health holds significant potential for enhancing maintenance strategies.
As Avangrid takes ownership of these AI systems, the utility envisions a trajectory of continuous improvement, boosting efficiency and cost-effectiveness over time. The convergence of AI and grid management highlights the utility’s dedication to delivering reliable services while advancing operational excellence in an evolving energy landscape.