MUMBAI, India, July 24 -- Intellectual Property India has published a patent application (202641085355 A) filed by Sri Eshwar College Of Engineering on July 11, 2026, for Self-Learning Digital Twin Architecture For Ai-Driven Health Monitoring And Adaptive Control Of Semiconductor Devices.
Inventors include Ms. K. Gowthami; Ms. J. Yashwandra; Mr. S. Aravind; and Ms. Minu Balakrishnan.
The application for the patent was published on July 17, 2026, under issue no. 29/2026.
Abstract: A semiconductor health monitoring system based on Al, digital twin and adaptive control is structured to automatically monitor device health, forecast degradation, and improve operational reliability in continuously-synchronized closed-loop architecture. Electrical, thermal, current, voltage, power, leakage current, and optional environmental parameters, are robotically transmitted from one or more sensor(s) related to a semiconductor device. The acquired data are synchronized, validated, filtered, and processed to develop a comprehensive feature representation for an intelligent health assessment.An artificial intelligence engine analyzes the extracted features to assess the health condition, identify degradation characteristics, estimate the stress levels, and produce predictive health information. Prediction results are continually synchronized with a digital twin that maintains a virtual representation of the semiconductor device in real-time. Synchronization is performed through a self-learning process that compares physical measurements and virtual predictions and incrementally updates model parameters for improved prediction accuracy during operation. A degradation-aware model selection approach enables the automatic activation of the appropriate digital twin, based on the failure mechanism identified, for condition-specific analysis. The remaining useful life of a component is predicted by combining the information provided by the AI algorithm with the behavior of the digital twin, in order to forecast future degradation, and by assessing the confidence of the predictions. Based on the synchronized health state and the development of the degradation and the remaining useful life (RUL) forecast an adaptive control arranges the optimal operation commands for the semiconductor. The execution feedback is incorporated into the learning process for the improvement of the quality of the forecast and the control in a closed-loop fashion. In this way the reliability is increased, the degradation is reduced, the quality of the forecast is improved, and the operation life of the semiconductor is extended without any intervention. In one exemplary embodiment, the method is utilized within electronic power circuits for electric vehicles to automatically identify degradation, predict imminent failures and then adjust the device operation to prevent degradation before the device ceases to operate successfully, which results in increased system safety, availability and service life.
Disclaimer: Curated by HT Syndication.
