Rate My Professor Daniel Polani

Professor Daniel Polani is Professor of Artificial Intelligence in the Department of Computer Science within the School of Physics, Engineering and Computer Science at the University of Hertfordshire. He serves as Head of the Adaptive Systems Research Group and is affiliated with the Centre for AI and Robotics Research. His research centers on understanding and imitating processes that enable animals and humans to make flexible decisions in complex environments and adapt gracefully. This includes incorporating learning abilities without compromising generalization and exploiting general principles of intelligent information processing independent of specialized solutions. Polani employs methods from Artificial Life, with a particular emphasis on Information Theory applied to areas such as sensor evolution, collective systems, and complex systems.

Key contributions include developing a principled formalism for the information-theoretic treatment of the perception-action loop grounded in information flow; an information-theoretic bottom-up approach to reconstructing world models in real-world robots; the concept of 'empowerment' as an information-theoretic universal utility function derived from agent embodiment; explorations of informational constraints on cognition and behavior; and informational conservation laws governing cognitive behavior. His work also examines information bookkeeping in cognitive agents to tackle challenges in distributed robotic designs and scenarios with limited information processing capacity. Polani has produced 185 research outputs, comprising 82 journal articles, 68 conference contributions, and more. Notable recent publications are 'SuPLE: Robot Learning with Lyapunov Rewards' (Nguyen, P., Polani, D., Tiomkin, S., 2025), 'Process empowerment for robust intrinsic motivation' (Tiomkin, S., Salge, C., Polani, D., 2025), 'Decentralized Traffic Flow Optimization Through Intrinsic Motivation' (Papala, H., Polani, D., Tiomkin, S., 2025), 'An Informational Parsimony Perspective on Symmetry-Based Structure Extraction' (Charvin, H., Volpi, N. C., Polani, D., 2024), and 'Dimensionality Reduction of Dynamics on Lie Groups via Structure-Aware Canonical Correlation Analysis' (Chung, W., Polani, D., Tiomkin, S., 2024). His research topics encompass empowerment for intrinsic motivation, perception-action loops, world model reconstruction for robotics, informational constraints in cognition, and applications in robotics and complex systems.