Embodied Intelligence: Where Control Meets Cognition in the Physical World
Embodied intelligence is rapidly becoming a unifying scientific frontier. Recent advances in learning, sensing, and computation increasingly require principled control, rigorous modeling, and systematic real-world validation to enable safe, robust, and generalizable physical autonomy. However, much of the most impactful work lies at the intersection of traditionally separate fields—such as control theory, robotics, artificial intelligence, neuroscience, and human-centered systems—where appropriate standards for evaluation and dissemination are often unclear. Embodied Intelligence is founded to provide a dedicated scholarly forum for this cross-disciplinary community. The journal promotes research that integrates theoretical depth with embodied grounding, and that accelerates reproducible progress from fundamental principles to deployed systems.
What We Publish
Embodied Intelligence welcomes original research articles, review articles, perspectives, commentaries, technology reports, system papers, case studies, and software/database notes in (but not limited to) the following areas.
1. Foundations: Modeling, Control, and Optimization for Embodied Agents
- Nonlinear, hybrid, and stochastic control for complex embodied systems and contact-rich interactions
- Optimal control, model predictive control (MPC), and real-time trajectory optimization
- System identification and learning-based modeling with guarantees of stability and robustness
- Safety, constraints, and formal verification (e.g., barrier functions, reachability, and certified control)
2. Learning for Embodiment: From Data to Skills with Guarantees
- Reinforcement learning and imitation learning for continuous control and interaction
- Sim-to-real transfer, domain adaptation, and generalization under distribution shift
- Representation learning for dynamics, contact modeling, and morphology-aware policies
- Safe and robust learning: uncertainty quantification, risk-sensitive design, and policy certification
3. Perception, State Estimation, and Sensorimotor Integration
- Multimodal perception (vision, tactile sensing, proprioception, and audio) for embodied interaction
- State estimation, simultaneous localization and mapping (SLAM), and contact-aware localization and manipulation
- World modeling: scene understanding, affordance learning, and physics-informed perception
- Closed-loop sensorimotor architectures that tightly integrate perception and control
4. Systems, Platforms, and Real-World Impact
- Novel robot designs and embodied hardware (soft robotics, dexterous manipulation, and bio-inspired systems)
- Human–robot interaction, shared autonomy, and assistive/medical/rehabilitation applications
- Swarm and multi-agent embodied intelligence, addressing coordination, resilience, and communication constraints
- Benchmarking, reproducibility, datasets, and field deployments in industrial, service, and extreme environments
