Foundations and Trends® in Robotics > Vol 12 > Issue 4

Foundations of Embodied Intelligence for Robotic Systems

By Fumiya Iida, University of Cambridge, UK, fi224@cam.ac.uk | Chapa Sirithunge, University of Cambridge, UK, csh66@cam.ac.uk | Perla Maiolino, University of Oxford, UK, perla.maiolino@eng.ox.ac.uk | Josie Hughes, Swiss Federal Institute of Technology, Lausanne, Switzerland, josie.hughes@epfl.ch

 
Suggested Citation
Fumiya Iida, Chapa Sirithunge, Perla Maiolino and Josie Hughes (2025), "Foundations of Embodied Intelligence for Robotic Systems", Foundations and TrendsĀ® in Robotics: Vol. 12: No. 4, pp 350-435. http://dx.doi.org/10.1561/2300000080

Publication Date: 19 Nov 2025
© 2025 F. Iida et al.
 
Subjects
Artificial intelligence in robotics,  Human-robot interaction,  Robot control,  Mechanisms and actuators,  Behavioral synthesis,  Behavioral, cognitive and neural learning
 

Free Preview:

Download extract

Share

Download article
In this article:
1. Introduction
2. Foundations of Embodied Intelligence
3. Competing Paradigms and Architectures of Embodied Intelligence
4. Discussions and Perspectives
Acknowledgements
References

Abstract

Embodied Intelligence (EI) provides a foundational framework for understanding intelligent and adaptive behaviors in both biological and artificial systems. Emerging from early philosophical and scientific discussions on mind-body dualism, EI emphasises the intricate relationship between physical bodies and information processing, which enables meaningful sensory-motor interactions with the environment. Its theories and technologies span multiple disciplines, including mechanical, electrical and control engineering, as well as computer science and material sciences, making it a complex field for robotics researchers to navigate. However, a comprehensive understanding of EI is essential to addressing key challenges and avoiding the redundancy of past discoveries. This article introduces fundamental design principles of intelligent adaptive systems, followed by key paradigms in autonomous agent architectures, such as feedback control, behavior-based approaches, mechanical and material intelligence and embodied social interaction. It explores insights from biology, physics, cognitive science and philosophy, focusing on emerging applications in adaptive systems, human-robot collaboration and bio-hybrid technologies.

DOI:10.1561/2300000080
ISBN: 978-1-63828-651-6
98 pp. $160.00
Buy E-book (.pdf)
Table of contents:
1. Introduction
2. Foundations of Embodied Intelligence
3. Competing Paradigms and Architectures of Embodied Intelligence
4. Discussions and Perspectives
Acknowledgements
References

Foundations of Embodied Intelligence for Robotic Systems

Embodied Intelligence (EI) provides a foundational framework for understanding intelligent and adaptive behaviors in both biological and artificial systems. Emerging from early philosophical and scientific discussions on mind-body dualism, EI emphasizes the intricate relationship between physical bodies and information processing, which enables meaningful sensory-motor interactions with the environment. Its theories and technologies span multiple disciplines, including mechanical, electrical and control engineering, as well as computer science and material sciences, making it a complex field for robotics researchers to navigate. However, a comprehensive understanding of EI is essential to addressing key challenges and avoiding the redundancy of past discoveries.

This monograph introduces fundamental design principles of intelligent adaptive systems, followed by key paradigms in autonomous agent architectures, such as feedback control, behavior-based approaches, mechanical and material intelligence and embodied social interaction. It explores insights from biology, physics, cognitive science and philosophy, focusing on emerging applications in adaptive systems, human-robot collaboration and bio-hybrid technologies.

 
ROB-080