Predictive information and explorative behavior of autonomous robots

Abstract.Measures of complexity are of immediate interest for the field of autonomous robots both as a means to classify the behavior and as an objective function for the autonomous development of robot behavior. In the present paper we consider predictive information in sensor space as a measure for the behavioral complexity of a two-wheel embodied robot moving in a rectangular arena with several obstacles. The mutual information (MI) between past and future sensor values is found empirically to have a maximum for a behavior which is both explorative and sensitive to the environment. This makes predictive information a prospective candidate as an objective function for the autonomous development of such behaviors. We derive theoretical expressions for the MI in order to obtain an explicit update rule for the gradient ascent dynamics. Interestingly, in the case of a linear or linearized model of the sensorimotor dynamics the structure of the learning rule derived depends only on the dynamical properties while the value of the MI influences only the learning rate. In this way the problem of the prohibitively large sampling times for information theoretic measures can be circumvented. This result can be generalized and may help to derive explicit learning rules from complexity theoretic measures.

• Publication year

  2008

• Venue

  The European Physical Journal B

• Publication date

  2008-04-24

• Fields of study

  Computer Science, Engineering

• Identifiers
  DOI 10.1140/EPJB/E2008-00175-0
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Methods for quantifying the informational structure of sensory and motor data

  2007cited by this paper
• Open Dynamics Engine を用いたスノーボードロボットシミュレータの開発

  2007cited by this paper
• Representations of Space and Time in the Maximization of Information Flow in the Perception-Action Loop

  2007cited by this paper
• Rocking Stamper and Jumping Snakes from a Dynamical Systems Approach to Artificial Life

  2006cited by this paper
• From Motor Babbling to Purposive Actions: Emerging Self-exploration in a Dynamical Systems Approach to Early Robot Development

  2006cited by this paper
• Let it roll - Emerging Sensorimotor Coordination in a Spherical Robot

  2006cited by this paper
• Elements of Information Theory (Wiley Series in Telecommunications and Signal Processing)

  2006cited by this paper
• Mapping Information Flow in Sensorimotor Networks

  2006cited by this paper
• Intrinsically Motivated Reinforcement Learning: A Promising Framework for Developmental Robot Learning

  2005cited by this paper
• The Playground Experiment: Task-Independent Development of a Curious Robot

  2005cited by this paper
• Completely Self-referential Optimal Reinforcement Learners

  2005cited by this paper
• Empowerment: a universal agent-centric measure of control

  2005cited by this paper
• Elements of Information Theory

  2005cited by this paper
• Developmental robotics: a survey

  2003cited by this paper
• Predictability , Complexity , and Learning

  2002cited by this paper
• True autonomy from self-organized adaptivity

  2002cited by this paper
• Self-organized acquisition of situated behaviors

  2001cited by this paper
• Autonomous Mental Development by Robots and Animals

  2001cited by this paper
• Predictability, Complexity, and Learning

  2000cited by this paper
• Homeokinesis - A new principle to back up evolution with learning

  1999cited by this paper
• Relative Information — What For?

  1990cited by this paper
• Inferring statistical complexity.

  1989cited by this paper
• Toward a quantitative theory of self-generated complexity

  1986cited by this paper
• Fokker-Planck Equation

  1984cited by this paper
• Time series analysis, forecasting and control

  1972cited by this paper

• Gradient-Free De Novo Learning

  2025cites this paper
• Active Inference and Intentional Behavior

  2025cites this paper
• Defining and finding lifelike entities with a lazy filter

  2025cites this paper
• From pixels to planning: scale-free active inference

  2025cites this paper
• Goals and the Structure of Experience

  2025cites this paper
• Information theory for complex systems scientists: What, why, and how

  2025cites this paper
• Why collective behaviours self-organize to criticality: a primer on information-theoretic and thermodynamic utility measures

  2024influential citation
• The many faces of action: Comment on "An active inference model of hierarchical action understanding, learning and imitation" by Proietti, Pezzulo, and Tessari.

  2023cites this paper
• Federated inference and belief sharing

  2023cites this paper
• Detecting Information Relays in Deep Neural Networks

  2023cites this paper
• Supervised structure learning

  2023cites this paper
• Selection for short-term empowerment accelerates the evolution of homeostatic neural cellular automata

  2023cites this paper
• Active Inference and Intentional Behaviour

  2023cites this paper
• Information Dynamics of Electric Field Intensity before and during the COVID-19 Pandemic

  2022cites this paper
• On the Criticality of Adaptive Boolean Network Robots

  2022cites this paper
• Empowered neural cellular automata

  2022cites this paper
• Systems Challenges for Trustworthy Embodied Systems

  2022cites this paper
• An Information-Motivated Exploration Agent to Locate Stationary Persons with Wireless Transmitters in Unknown Environments

  2021cites this paper
• A Dynamic Programming Algorithm for Finding an Optimal Sequence of Informative Measurements

  2021cites this paper
• A dynamic programming algorithm for informative measurements and near-optimal path-planning

  2021cites this paper
• Can Being Aware of the Illusion of Self Augment an Agent's Affordances: Integrating Buddhist Philosophy, Cognitive Science, and Artificial Life

  2021cites this paper
• Space Mining Robot Prototype for NASA Robotic Mining Competition Utilizing Systems Engineering Principles

  2021cites this paper
• Developing a Mining Robot for Mars Exploitation: NASA Robotics Mining Competition (RMC)

  2020cites this paper
• Goal-Directed Empowerment: Combining Intrinsic Motivation and Task-Oriented Behavior

  2020cites this paper
• Localized active learning of Gaussian process state space models

  2020cites this paper
• Tutorial Review of Bio-Inspired Approaches to Robotic Manipulation for Space Debris Salvage

  2020cites this paper
• Online adaptation in robots as biological development provides phenotypic plasticity

  2020cites this paper
• Empowerment and Relevant Goal Information as Alternatives to Graph-Theoretic Centrality for Navigational Decision Making

  2020cites this paper
• Human perception of intrinsically motivated autonomy in human-robot interaction

  2020influential citation
• Conditioning in Tropical Probability Theory

  2019cites this paper
• Tropical diagrams of probability spaces

  2019cites this paper
• Complexity Measures: Open Questions and Novel Opportunities in the Automatic Design and Analysis of Robot Swarms

  2019cites this paper
• Evaluation of Internal Models in Autonomous Learning

  2019cites this paper
• Persistent random motion with maximally correlated fluctuations.

  2019cites this paper
• Consciousness in Humanoid Robots

  2019cites this paper
• Structural transition in the collective behavior of cognitive agents

  2019cites this paper
• Factorized Mutual Information Maximization

  2019cites this paper
• Active inference: building a new bridge between control theory and embodied cognitive science

  2019cites this paper
• Fractals2019: Combinatorial Optimisation with Dynamic Constraint Annealing

  2019cites this paper
• On Markov blankets and hierarchical self-organisation

  2019cites this paper
• The Evolution of Neuroplasticity and the Effect on Integrated Information

  2019cites this paper
• Intrinsically Motivated Autonomy in Human-Robot Interaction: Human Perception of Predictive Information in Robots

  2019cites this paper
• Tropical diagrams of probability spaces

  2019cites this paper
• Arrow Contraction and Expansion in Tropical Diagrams

  2019cites this paper
• Expanding the Active Inference Landscape: More Intrinsic Motivations in the Perception-Action Loop

  2018cites this paper
• Multiscale information storage of linear long-range correlated stochastic processes.

  2018cites this paper
• Asymptotic dependency structure of multiple signals

  2018cites this paper
• Framework for development algorithmics swarm elligence to applications in supply warehouse

  2018cites this paper
• Free energy , empowerment , and predictive information compared

  2018cites this paper
• Happily entangled: prediction, emotion, and the embodied mind

  2018cites this paper
• Self-motivated composition of strategic action policies

  2018cites this paper
• Entangled predictive brain: emotion, prediction and embodied cognition

  2018cites this paper
• Bio-inspired self-organising multi-robot pattern formation: A review

  2017cites this paper
• RoboCup 2017: Robot World Cup XXI

  2017cites this paper
• Tropical Limits of Probability Spaces, Part I: The Intrinsic Kolmogorov-Sinai Distance and the Asymptotic Equipartition Property for Configurations

  2017cites this paper
• Towards controlled self-organized behavioral exploration

  2017cites this paper
• Evolution of Decision-Making Systems

  2017cites this paper
• Addressing the "Why?" in Computational Creativity: A Non-Anthropocentric, Minimal Model of Intentional Creative Agency

  2017cites this paper
• Biological Self-organisation and Markov blankets

  2017cites this paper
• Computational Foundations of Natural Intelligence

  2017cites this paper
• Complexity Measures in Automatic Design of Robot Swarms: An Exploratory Study

  2017cites this paper
• Empowerment As Replacement for the Three Laws of Robotics

  2017cites this paper
• Control Capacity of Partially Observable Dynamic Systems in Continuous Time

  2017cites this paper
• RoboCup 2D Soccer Simulation League: Evaluation Challenges

  2017cites this paper
• Entropy balance and information processing in bipartite and nonbipartite composite systems

  2017cites this paper
• Measuring motion significance and motion complexity

  2017cites this paper
• Empowerment as a Generic Utility Function for Agents in a Simple Team Sport Simulation

  2017cites this paper
• Self-Organized Behavior Generation for Musculoskeletal Robots

  2017cites this paper
• Closed-loop Robots Driven by Short-Term Synaptic Plasticity: Emergent Explorative vs. Limit-Cycle Locomotion

  2016cites this paper
• Transposing Simulated Self-organizing Robots into Reality using the PLUG&LEARN Architecture

  2016cites this paper
• In Search for the Neural Mechanisms of Individual Development: Behavior-Driven Differential Hebbian Learning

  2016cites this paper
• Ich bin – also denke ich

  2016cites this paper
• The influence of active vision on the exoskeleton of intelligent agents

  2016cites this paper
• Dynamical self-consistency leads to behavioral development and emergent social interactions in robots

  2016cites this paper
• Novel plasticity rule can explain the development of sensorimotor intelligence

  2015cites this paper
• Information-theoretic neuro-correlates boost evolution of cognitive systems

  2015cites this paper
• The Intrinsic Cause-Effect Power of Discrete Dynamical Systems - From Elementary Cellular Automata to Adapting Animats

  2015influential citation
• The Sensorimotor Loop as a Dynamical System: How Regular Motion Primitives May Emerge from Self-Organized Limit Cycles

  2015cites this paper
• Information transfer and causality in the sensorimotor loop

  2015cites this paper
• Quantifying Emergent Behavior of Autonomous Robots

  2015influential citation
• Information Geometry on Complexity and Stochastic Interaction

  2015cites this paper
• Digested Information as an Information Theoretic Motivation for Social Interaction

  2015cites this paper
• Principles of Neural Information Processing

  2015cites this paper
• On the Cross-Disciplinary Nature of Guided Self-Organisation

  2014cites this paper
• A review of automation in manufacturing illustrated by a case study on mixed-mode hot forging

  2014cites this paper
• Information-based learning by agents in unbounded state spaces

  2014cites this paper
• Artificial Intelligence Algorithm in Mobile Robots

  2014cites this paper
• JIDT: An Information-Theoretic Toolkit for Studying the Dynamics of Complex Systems

  2014cites this paper
• Evolution of Integrated Causal Structures in Animats Exposed to Environments of Increasing Complexity

  2014influential citation
• General Self-Motivation and Strategy Identification: Case Studies Based on Sokoban and Pac-Man

  2014cites this paper
• Grand Challenges for Computational Intelligence

  2014influential citation
• Guided Self-Organization in a Dynamic Embodied System Based on Attractor Selection Mechanism

  2014cites this paper
• Cognitive Dynamics: From Attractors to Active Inference

  2014cites this paper
• Ontogenesis of Agency in Machines : A Multidisciplinary Review

  2014cites this paper
• On the Causal Structure of the Sensorimotor Loop

  2014cites this paper
• Changing the Environment Based on Empowerment as Intrinsic Motivation

  2014cites this paper
• Generating Functionals for Computational Intelligence: The Fisher Information as an Objective Function for Self-Limiting Hebbian Learning Rules

  2014cites this paper
• Information driven self-organization of agents and agent collectives

  2014cites this paper
• On the Role of Embodiment for Self-Organizing Robots: Behavior As Broken Symmetry

  2014cites this paper
• Causal Models over Infinite Graphs and their Application to the Sensorimotor Loop - General Stochastic Aspects and Gradient Methods for Optimal Control

  2014cites this paper