Research Terms
Artificial Intelligence Renewable Energy Resources Water Power Electrical Networks Power Systems
Industries
Biofuels Smart Grid Water Renewable Energy
Smart grid systems face significant challenges with power transformer maintenance due to the complex data generated and the critical role transformers play in grid reliability. Current solutions, such as reactive maintenance and scheduled inspections, often fall short of managing this complexity. Reactive maintenance is costly and introduces delays as faults are often only addressed after they've led to system issues. Scheduled inspections, while proactive, are time-intensive and do not capture real-time fault conditions. These drawbacks lead to increased operational costs, extended downtimes, and reduced grid reliability, emphasizing the need for a predictive and real-time fault management solution.
Researchers at Florida Atlantic University have developed PowerGPT, a cutting-edge artificial intelligence system that leverages advanced machine learning, natural language processing, and cloud-based frameworks for big data analysis and management in smart grid systems. The system provides real-time, predictive diagnostics for power transformers, enabling proactive management of grid health and reducing costly downtime associated with reactive maintenance. Unlike traditional methods, PowerGPT integrates a high-accuracy classification model into a user-friendly, conversational interface that can analyze vast datasets quickly and effectively, identifying and categorizing faults with 97% accuracy. Currently, the innovation is at the proof-of-concept stage, with successful demonstrations of its diagnostic capabilities, and development is underway to expand its real-world application potential.
FAU seeks to advance this innovation into the marketplace through licensing or development partnerships.
A ChatGPT-like Solution for Power Transformer Condition Monitoring
Surgical planning and post-operative rehabilitation are hindered by fragmented workflows, limited personalization, and inefficient use of medical data. Current solutions rely heavily on manual interpretation of 2D CT and MRI images, standalone surgical planning software, generic implant templates, and disconnected rehabilitation tools. These approaches are time-intensive, prone to variability, and poorly suited for predicting patient-specific risks or outcomes. As a result, surgeons face longer planning cycles and uncertainty in implant selection, while patients often experience suboptimal outcomes, reduced engagement, and slower recovery.
Researchers at Florida Atlantic University have developed an AI-driven platform that integrates surgical planning, implant design, and post-operative rehabilitation into a unified, patient-specific workflow. The innovation automatically converts standard medical imaging into 3D anatomical models, applies machine learning to compare cases against historical outcomes, and supports personalized surgical and rehabilitation decisions. By unifying these capabilities in a single system, the platform improves precision, reduces manual effort, and enhances patient engagement compared to existing point solutions. The technology is currently in active development, with ongoing efforts focused on prototype refinement and validation.
FAU seeks to advance this innovation into the marketplace through licensing or development partnerships.
