Stefano Nolfi

Laboratory of Autonomous Robots and Artificial Life

Institute of Cognitive Sciences and Technologies (ISTC)

National Research Council (CNR)

Roma, Italy. email: stefano.nolfi (at) istc.cnr.it

My photo

Publications


Bio & CV

My Projects

Behavioral and Cognitive Robotics

Behavioral and Cognitive Robotics: An Adaptive Perspective. This open access book, which is targeted toward researchers, PhD and Master students with an interest in machine learning and robotics: (i) introduces autonomous robots, evolutionary algorithms, reinforcement learning algorithms, and learning by demonstration methods, (ii) uses concrete experiments to illustrate the fundamental aspects of embodied intelligence, (iii) provides theoretical and practical knowledge, including tutorials and exercises, and (iv) provides an integrated review of recent research in this area carried within partially separated research communities.

Evolution of Communication and Language in Embodied Agents

Evolution of Communication and Language. I studied how a population of initially non-communicating robots can evolve a communication system from scratch, how such communication system can evolve and complexify, and how robots can acquire the ability to comprehend human language. See in particular the book I edited with Springer Verlag (left figure) and the following articles (keep the mouse over the links to see the corresponding reference):

  1. Communication and language
  2. Emergence of linguistic compositionality in evolving robots
  3. Emergence of communication and language in evolving robots
  4. Evolution of implicit and explicit communication
  5. Emerge of communication in evolving robots

Evolutionary Robotics

Evolutionary Robotics. This book entitled Evolutionary Robotics: The Biology, Intelligence and Technology of Self-Organized Machines published with MIT Press in 2000 became the reference text for the research community. It sold more than 2500 copies and collected a similar number citations. For a more recent review of evolutionary robotics, see: Evolutionary Robotics. Keep the mouse over the links to see the corresponding reference.

Cognition in Adaptive Robots

Cognition in Adaptive Robots. I investigated the evolution of cognitive skills in adapting robots. See in particular the following articles (keep the mouse over the links to see the corresponding reference):

  1. Learning of features for control
  2. Selective attention
  3. Categorization in human and artificial agents
  4. Development of reaching and grasping capabilities
  5. Selective attention and behavior arbitration
  6. Evolution and usage of a forward model which allows to cope with sensory deprivation
  7. Language and action compositionality
  8. Active categorical perception
  9. Spatial representation and the coupling of the robot's internal and external dynamics
  10. Learning to perceive the world as articulated
Neuroevolution

Neuroevolution I investigated the efficacy of different neuro-evolutionary methods and the factors that promote the evolution of solutions which are effective and robust to variations. See in particular the following articles (keep the mouse over the links to see the corresponding reference):

  1. Automated curriculum learning
  2. Robust optimization through neuroevolution
  3. Utility of moderate environmental variation
  4. Maximizing adaptive power in neuroevolution
Competitive Co-Evolution

Competitive Co-evolution. I studied the dynamic of competitive co-evolutionary processes and how the evolution of competing robots can give rise to long-term progress. These researches are reported in the following articles (keep the mouse over the links to see the corresponding reference):

  1. Long-term progress and progressive complexification in co-evolving robots
  2. Co-evolving predator and prey robots
  3. Arm races in artificial evolution
  4. Competition in evolutionary robotics
  5. Adaptive behavior in competing co-evolving robots
Evolution of Cooperative Behaviors, Self-organization and Swarm Robotics

Evolution of Cooperative Behaviors, Self-Organization and Swarm Robotics. I studied the evolution of coordinated and cooperative behaviors in groups of robots and the emergence of self-organization in groups of evolving robots. These researches are reported in the following articles (keep the mouse over the links to see the corresponding reference):

  1. Evolutionary Swarm Robotics
  2. Evolution of reciprocity
  3. Self-organized path formation in evolving swarmbots
  4. Self-organising sync in a robotic swarm
  5. Sythesizing self-organized behaviors by maximizing mutual information
  6. Coordinated motion in groups of physically connected robots
  7. Coordinated hole avoidance
Enactive Intelligence

Enactive Intelligence. I studied how behavioral and cognitive skills can emerge from the dynamical interaction between an agent and its environment. Moreover I analyzed the importance of considering behavior and cognition as dynamical systems with multi-level and multi-scale organizations. See in particular the following articles (keep the mouse over the links to see the corresponding reference):

  1. Situatedness
  2. Making the environment an informative place
  3. Cognitive offloading
  4. Behavioral Plasticity
  5. Behavior and cognition as a complex adaptive system
  6. Behaviour as a complex adaptive system
  7. Power and limits of reactive agents

Evolvability, modularity, learning & evolution I investigated the role of phenotypic organization in evolution and the possibility to combine evolution and learning. See in particular the following articles (keep the mouse over the links to see the corresponding reference):

  1. Enhancing Cartesian genetic programming through preferential selection of larger solutions
  2. Robustness, evolvability, and phenotypic complexity
  3. Robustness to faults promotes evolvability
  4. Integrating learning by experience and demonstration
  5. Duplication of modules facilitates the evolution of functional specialization
  6. How learning and evolution interact
  7. Learning and Evolution
  8. Learning to adapt to changing environments
  9. Learning and evolution in neural networks