Andreas Schaarschmidt
Papers
1
Total Citations
18
H-Index
1
About
Andreas Schaarschmidt is a leading researcher at the intersection of robotics and machine learning, with a primary focus on model-based reinforcement learning (MBRL) and safe, uncertainty-aware control. His most impactful work, "Uncertainty-Aware Contact-Safe Model-Based Reinforcement Learning" (2021, 18 citations), addresses a critical challenge in deploying MBRL on physical robots: ensuring safety during contact-rich tasks. Schaarschmidt’s key contribution is a framework that explicitly models predictive uncertainty to avoid dangerous collisions and unstable interactions while the robot is still learning, enabling more reliable and sample-efficient policy acquisition. This work bridges the gap between data-driven learning and real-world safety constraints, making it foundational for researchers developing autonomous systems that must operate safely in unstructured environments. Beyond this, his broader research advances robust decision-making under uncertainty, with applications in manipulation and locomotion. Schaarschmidt’s work is essential reading for anyone seeking to understand how to build learning agents that are not only intelligent but also inherently cautious and contact-aware.
Research Focus
Key Achievements
Top Papers
- 1Uncertainty-Aware Contact-Safe Model-Based Reinforcement Learning18 citations · 2021