Xingjian (Ken) Yan

Xingjian (Ken) Yan

PhD Student, MIT Department of Earth, Atmospheric, and Planetary Sciences

Computer Vision on High-Voltage Energized Equipment

Maintaining a reliable electricity supply for metropolitan areas requires substantial effort. Our family business, PMDT, is exploring trailblazing methods to combine decades of expertise with modern computer vision technology to enhance the diagnosis and prevention of equipment failures from early stages.

My cousins Xiaolan (Angelina) Huang, Feiyan (Jennifer) Ma and I partnered with Vitally AI to develop a comprehensive model that integrates equipment detection, automated temperature anomaly warnings, equipment failure diagnoses, and solution generation.

With 87% precision in object detection and 100% precision in failure diagnoses, our closed-source algorithm has significantly reduced the manpower needed to monitor crucial substation equipment and issue early warnings. This achievement is due to PMDT's industry-leading database of \(O(10^6)\) records from both natural light and infrared sources.

For business inquiries, please contact PMDT.

Due to regulatory and disclosure requirements, I am only attaching a schematic object identification figure here.