Active Inference in Robotics and Artificial Agents: Survey and Challenges

TL;DR

This survey reviews active inference as a promising framework for robotics and artificial agents, highlighting recent advances, experimental results, and discussing its benefits and challenges in state-estimation, control, and learning.

Contribution

It provides a comprehensive overview of active inference applications in robotics, connecting it with other frameworks and discussing future challenges and benefits.

Findings

Active inference enables robust adaptation and generalization in robotics.

Experimental results demonstrate effective control and learning capabilities.

The framework offers a biologically plausible approach to decision-making.

Abstract

Active inference is a mathematical framework which originated in computational neuroscience as a theory of how the brain implements action, perception and learning. Recently, it has been shown to be a promising approach to the problems of state-estimation and control under uncertainty, as well as a foundation for the construction of goal-driven behaviours in robotics and artificial agents in general. Here, we review the state-of-the-art theory and implementations of active inference for state-estimation, control, planning and learning; describing current achievements with a particular focus on robotics. We showcase relevant experiments that illustrate its potential in terms of adaptation, generalization and robustness. Furthermore, we connect this approach with other frameworks and discuss its expected benefits and challenges: a unified framework with functional biological plausibility using variational Bayesian inference.