This is Yihao Cai, AKA Charles. I’m currently a PhD student in the Department of Electrical and Computer Engineering at Wayne State University (WSU). My research primarily focuses on Physics-AI, Cyber Physical Systems (CPS), AI Safety, and their applications in Robotics. My goal is to dedicate myself to the advancement of the AI field within the aim of enhancing the safety and intelligence of robots.

Physics-AI

In contrast to pure AI, Physics-AI emphasizes integrating physical knowledge with AI agents. More specifically, it could be divided into two subfields: Physics for AI and AI for Physics:

Physics for AI: In this context, Physics refers to domain-specific knowledge acquired through human understanding. This knowledge can be integrated with AI to enhance performance. For instance, pre-existing model knowledge or principles can be embedded into neural networks during training to achieve faster convergence. Additionally, since all data-driven approaches are susceptible to environmental changes, deploying AI models in safety-critical systems can be challenging. In such cases, incorporating established physical laws alongside AI can help ensure safety and increase reliability.

AI for Physics: Owing to the notorious AI Hallucination issue, models generated by AI may not always adhere strictly to the physical laws governing the real world. To address this, prior physical laws can be encoded into models to ensure they respect these principles. Moreover, given AI’s strengths in function approximation, there is potential to use AI to discover latent general physical laws that apply to various real-life scenarios. These laws are often represented in the form of partial differential equations (PDEs). The term PINNs refers to this approach.

I believe that Physics-AI is a significant and timely research topic within the rapidly evolving field of AI, and I am eager to make meaningful contributions to it throughout my PhD career.

News

Hobbies

League of Legends | CS Online | Basketball | Soccer | Chinese Chess | Hacking | Sleeping~zzz

For more info

More info about configuring academicpages can be found in the guide. The guides for the Minimal Mistakes theme (which this theme was forked from) might also be helpful.