Embodied AI
Developing intelligent systems that perform complex tasks through understanding and engaging with physical and virtual environments
What is embodied AI?
Embodied AI systems interact with and learn from the physical world through sensory inputs and actions, using robotic or virtual tools to perceive, navigate, and interact with their environment.
Our mission is to accelerate and advance research to develop AI agents that can:
- Learn by interacting with the real world
- Adapt to dynamic environments through trial and error
- Transfer knowledge between virtual and physical spaces
- Translate perceptions into actions for completing everyday tasks like opening doors, picking up objects, or navigating around obstacles
Robots are typically machines equipped with actuators designed to execute specific tasks. Embodied AI is inherently interdisciplinary, involving robotic control, reinforcement learning, spatial awareness, human-robot interaction, reasoning, and more. Given this complexity, no single organization can cover all aspects of its development alone. We look forward to collaborating with local industry, academia, and institutions across the globe, leveraging their expertise alongside our strengths in AI to advance the field responsibly.
“We are working to advance embodied AI systems capable of interpreting and acting upon natural expressions of human intent. These systems will interact with the same semi-structured environments in which people live and work.”
– Ashley Llorens, CVP & Managing Director Microsoft Research Accelerator
Advancing intelligent embodiments
We are enabling AI systems that integrate perception, reasoning, and control, enabling adaptive and autonomous interaction in dynamic environments.
StarTrack Scholars Program accelerates embodied AI and large robotics models
The Microsoft Research Asia StarTrack Scholars Program is actively investing in a collaborative effort to explore Embodied AI and Large Action Models. We believe that the research on embodied AI will build a solid foundation and promising development prospect for robot intelligence, benefit human society.
