Papers
8
Total Citations
49
H-Index
4
About
Matthieu Geist is a robotics and autonomous systems researcher whose work spans two interconnected domains: ultrasonic-based robotic inspection of metal structures and deep reinforcement learning for autonomous control systems. His most influential contributions lie in developing sophisticated SLAM (Simultaneous Localization and Mapping) frameworks that exploit ultrasonic guided waves to enable robots to inspect large metal structures such as storage tanks and ship hulls — work that has garnered up to 15 citations and represents a meaningful advance in non-destructive evaluation robotics. Alongside this, Geist has made notable strides in applying deep reinforcement learning to complex control challenges, including multicopter hovering, floating platform navigation for microgravity emulation, and continuous control architectures, reflecting a consistent interest in bridging simulation and real-world deployment. His 2019 benchmark of neural network architectures for system identification — evaluating over three hundred models — demonstrates a rigorous, empirical approach to machine learning methodology, further complemented by his work on importance sampling for handling noisy, unbalanced datasets. Collectively, Geist's research equips autonomous systems with smarter sensing, mapping, and learning capabilities, making him a noteworthy contributor at the intersection of robotics, signal processing, and machine learning.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5A Comprehensive Benchmark of Neural Networks for System Identification4 citations · 2019
- 6
- 7Importance Sampling for Deep System Identification3 citations · 2019
- 8