Konstantin M. Seiler

The University of Queensland, The University of Sydney, University of Technology Sydney

Papers

Total Citations

121

H-Index

About

Konstantin M. Seiler’s research lies at the intersection of robotics, planning under uncertainty, and multi-agent systems, with a particular focus on enabling agile, autonomous decision-making in complex environments. His most influential work introduces an online, approximate solver for Partially Observable Markov Decision Processes (POMDPs) with continuous action spaces—a critical advancement for robotics where precise motion planning must account for sensor and actuation noise. This paper has garnered 74 citations, reflecting its impact on the field. Seiler has also made notable contributions to corrective motion planning by leveraging Lie group symmetries, a mathematical framework that allows for fast, elegant path corrections without explicit control strategies. This work, cited 34 times, is especially relevant for locomotion and other robotic systems where symmetry can be exploited for efficiency. More recently, Seiler has advanced multi-agent planning with the introduction of multi-horizon Monte Carlo tree search (MH-MCTS), the first framework to integrate hierarchical, decentralized planning across multiple agents and time horizons. His research consistently pushes the boundaries of how robots can reason and act under uncertainty, making him a key figure in modern autonomous systems.