Daniel Kohlsdorf
Papers
2
Total Citations
19
H-Index
2
About
Daniel Kohlsdorf is a researcher whose work sits at the intersection of robotics, artificial intelligence, and interactive simulation, with a core focus on enabling robots to anticipate and recover from failures. His major contributions center on developing predictive models that allow robots to learn the likely outcomes of their actions before they occur. By leveraging interactive physics-based simulations, Kohlsdorf demonstrated how robots can collect realistic training data to forecast task failures—such as a spatula missing a pancake during a flip—and proactively replan their actions to avoid errors. His 2015 paper, “Learning action failure models from interactive physics-based simulations,” which has garnered 12 citations, and his 2014 work, “Learning task outcome prediction for robot control from interactive environments,” with 7 citations, are foundational to this approach. These contributions are particularly notable for advancing “action-aware” robotics, equipping machines with the common-sense reasoning needed for complex, real-world tasks like cooking. Kohlsdorf’s research bridges the gap between simulation and physical robot control, offering a scalable pathway toward more robust and autonomous robotic systems.
Research Focus
Key Achievements
Top Papers
- 1Learning action failure models from interactive physics-based simulations12 citations · 2015
- 2