Papers
21
Total Citations
213
H-Index
9
About
Daniel Nikovski is a robotics and machine learning researcher whose work sits at the intersection of reinforcement learning, probabilistic modeling, and robotic control. His research has made substantial contributions to some of the field's most challenging problems: enabling robots to learn and navigate dynamic environments without complete system knowledge, and developing intelligent controllers for precision assembly tasks. Nikovski's earlier work laid important theoretical groundwork through probabilistic and POMDP-based models for robot navigation, with his 2002 thesis on state-aggregation algorithms for learning probabilistic representations earning continued recognition. His more recent research has pushed model-based reinforcement learning forward, notably developing derivative-free learning frameworks using Gaussian Process Regression for systems where only positional measurements are available — a critical real-world constraint addressed in his 2020 paper (16 citations). His contributions to robotic assembly are particularly impactful, spanning anomaly detection during insertion tasks, sim-to-real transfer learning, and imitation-based compliance controllers that enable generalizable human-robot collaboration. These works collectively address the gap between laboratory performance and real-world deployment. With citations accumulating across a focused and coherent research portfolio, Nikovski's scholarship has meaningfully advanced both the theory and practice of intelligent, adaptive robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9Design of Adaptive Compliance Controllers for Safe Robotic Assembly10 citations · 2023
- 10Imitation and Supervised Learning of Compliance for Robotic Assembly9 citations · 2022