Eivind Samuelsen
Papers
10
Total Citations
117
H-Index
7
About
Eivind Samuelsen is a researcher in evolutionary robotics, focusing on the automatic design of robot morphologies and control systems using evolutionary algorithms. His work addresses the critical challenge of the "reality gap"—the discrepancy between simulated and real-world robot performance—and explores methods to bridge it. Samuelsen’s major contributions include developing a two-phase approach to overcome initial convergence in multi-objective evolution, which enhances the co-optimization of robot body and brain, and introducing dynamic mutation in MAP-Elites to improve robotic repertoire generation. He has also proposed bioinspired representations, such as a Hox gene-inspired generative encoding for evolving limbed morphologies, and distance measures for morphological diversification. His research demonstrates real-world applicability, with studies on adapting robot morphology and control to hardware limitations and using obstacles to promote robust gaits. With over 100 citations across his top papers, Samuelsen’s work has advanced the field by enabling more efficient exploration of complex search spaces and improving the transfer of evolved designs from simulation to reality. His notable achievements include pioneering the use of memetic algorithms for robot control evolution and adaptation, making his research highly relevant for students and researchers interested in autonomous robotics, evolutionary computation, and embodied intelligence.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Dynamic mutation in MAP-Elites for robotic repertoire generation16 citations · 2018
- 4
- 5Memetic robot control evolution and adaption to reality15 citations · 2016
- 6A hox gene inspired generative approach to evolving robot morphology12 citations · 2013
- 7
- 8
- 9Multi-objective Analysis of MAP-Elites Performance3 citations · 2018
- 10On Restricting Real-Valued Genotypes in Evolutionary Algorithms2 citations · 2021