Bio-embodied Physical AI Research Group
KEYWORDKeyword
OBJECTIVE Objectives
○ To recruit and nurture outstanding domestic and international postdoctoral researchers in AI S&T, establishing a specialized research workforce tailored to strategic industries in the Daegu region.
○ To foster a practical collaborative ecosystem through joint research with Institutes of Science and Technology, universities, and industry partners.
○ To secure national strategic technologies through mid-to-long-term convergence research, while strengthening regional technological self-reliance and global competitiveness.
INTRODUCTION Director's Message
"Designing the Future of Human-Centered Intelligence through Trust and Convergence."
Greetings.
I am Young-Sik Kim, Director of the Research Group for Bio-Embodied Physical AI.
Artificial intelligence has now evolved beyond simple algorithms into embodied intelligence that understands the human body and behavior, collaborating with people in real time.
Our research group stands at the forefront of this transformation, integrating symbiotic AI, ultra-lightweight on-device AI, security-embedded AI, and biosignal-based NPU into a single, integrated research framework.
To this end, we closely link each technological domain—from biosignal-based multimodal foundation models and real-time lightweight inference to federated learning-based privacy protection and low-power, high-precision AI semiconductor design—through shared data and testbed infrastructure, thereby simultaneously enhancing the trustworthiness and practicality of human-centered intelligent systems.
The Research Group for Bio-Embodied Physical AI also presents a new model in research operations alongside technological innovation.
Through a multi-mentor system in which InnoCORE Fellows are connected to faculty members from two or more institutions, we systematically cultivate Fusion-Minded Co-Researchers who combine AI competencies with multidisciplinary S&T expertise.
By consolidating the specialized capabilities of the four Institutes of Science and Technology and Seoul National University around DGIST, and by expanding an open collaborative ecosystem with domestic and international industry partners, we will build the foundation for Korea to emerge as a global leader in the field of hyper-intelligent, trustworthy embedded AI converging robotics and bio-technologies.
Thank you.
IMPACT Impact
Strengthening the Industrial Ecosystem and Competitiveness
The trust-enhanced, symbiotic, bio-embodied AI technologies developed through this research will provide a sustained and stable supply of core talent to the AI-driven industrial ecosystem, effectively strengthening the technological self-reliance and competitiveness of regional enterprises.
Virtuous Cycle of Technological Upgrading and Commercialization
By channeling intensive postdoc-driven research outcomes into prototyping and technology transfer across advanced domains—including affective AI, on-device inference, highly stretchable multimodal sensors, and security-embedded AI—a virtuous cycle will be established that drives technological upgrading and next-generation commercialization models for regional industries.
Leap Toward a Global Convergence Research Hub
Through linkages among DGIST, the Suseong Alpha City AI startup infrastructure, and global partner institutions, the research group will promote the demonstration and dissemination of citizen-facing AI solutions such as emotionally intelligent robots and digital therapeutics, laying the groundwork for securing national strategic technologies and establishing a global AI convergence research hub.
TALENT Ideal Talent
Fusion-Minded
Convergence-oriented talent who think holistically across diverse technological domains—including AI, bio, robotics, semiconductors, and security—transcending the boundaries of any single institution or discipline.
Trustworthy & Proactive
Self-driven researchers who internalize human-centered trustworthiness and security as core values in AI system design, independently define research problems, and conduct hands-on validation in on-device and real-robot environments.
Globally Collaborative
Responsible, team-oriented talent who contribute to a collaborative research culture that breaks down barriers among Institutes of Science and Technology, openly engage with global academic networks and industry partners, and produce world-class research outcomes.