Machines that sense, reason, and act
Physical intelligence—the integration of sensing, reasoning, and actuation within physical systems—is fundamentally redefining how machines interact with and operate within the real world.
Unlike purely digital or cloud-based artificial intelligence, physical intelligence requires embodied systems to perceive complex environments, adapt to dynamic conditions, and execute precise physical tasks. From autonomous robotics and smart manufacturing to bio-hybrid systems and intelligent prosthetics, the convergence of mind and machine is driving breakthroughs across economic, industrial, and societal dimensions. As a result, physical intelligence has emerged as a crucial intersectional field, drawing expertise not only from robotics, mechanical engineering, and computer science, but also from material science, neuroscience, cognitive psychology, and control systems.
The landscape of intelligent machines is undergoing a profound shift as we move from rigid automation toward truly adaptive, autonomous behavior. These emerging systems promise to revolutionize industrial productivity, healthcare delivery, and environmental monitoring. However, realizing this potential requires overcoming steep technical hurdles in hardware-software co-design, real-time learning, safety guarantees, energy efficiency, and environmental resilience—challenges that cannot be solved in isolation and demand deeply integrated, interdisciplinary approaches.
A forum for the whole field
The second IEEE workshop on Physical Intelligence: Systems and Applications (PISA) serves as the premier global forum dedicated to advancing the state of the art in embodied AI and intelligent physical systems. IEEE PISA provides a vital platform for researchers, engineers, scientists, and industry innovators to exchange pioneering ideas, showcase technical breakthroughs, and map out future trajectories for physically intelligent systems.
Call for papers
IEEE PISA invites submissions of high-quality technical papers addressing foundational theory, novel architectures, and emerging applications in physical intelligence. We welcome original contributions on a wide range of topics, including but not limited to:
- Embodied AI frameworks
- Soft and bio-inspired robotics
- Multi-modal perception
- Adaptive control algorithms
- Smart materials
- Human-robot interaction
- Neuromorphic computing for edge systems
- Real-world deployment case studies
- Vision-language-action models and robot foundation models
- Imitation learning and learning from demonstration
- Reinforcement learning and sim-to-real transfer
- Dexterous and contact-rich manipulation
- Humanoid and legged locomotion
- SLAM, 3D mapping, and scene understanding
- Tactile sensing and haptics
- Safe, verifiable, and certifiable autonomy
Instructions for authors
All submissions must be made electronically in PDF format. Manuscripts should be prepared in English and formatted to ensure proper display in standard PDF readers. Each paper may contain up to 6 pages of single-spaced content, including all figures and tables but excluding the reference list. Papers must use 8.5" × 11" pages, 10-point font with 12-point line spacing, and a two-column layout.
Papers must present original work that is not simultaneously submitted to or being reviewed by another venue. All accepted papers will appear in the IEEE MOST 2027 workshop proceedings. For each accepted paper, at least one author is required to register for the workshop.
The submission link will be announced shortly.
Workshop Organizers
Program Chair
