Papers
19
Total Citations
603
H-Index
9
About
Daniel Kappler is a leading researcher at the intersection of robotic manipulation, computer vision, and machine learning, with a focus on enabling robots to operate intelligently in unstructured, real-world environments. His work spans open-vocabulary scene understanding, reactive motion generation, and imitation learning. Kappler’s most impactful contribution is NLMap, an open-vocabulary, queryable scene representation that bridges large language models with real-world robotic task planning (125 citations). He has also advanced real-time perception tightly integrated with reactive motion generation for grasping under uncertainty (107 citations), and pioneered zero-shot task generalization through the BC-Z imitation learning framework (89 citations). Kappler developed the OpenGRASP benchmarking suite (49 citations), a standard environment for comparative grasping analysis, and introduced Riemannian Motion Policies (49 citations), a modular mathematical framework for motion generation. His work on visual attention for object search (56 citations) and pixel-wise joint angle regression for robot arm pose estimation (36 citations) further demonstrates his breadth. With over 500 total citations and a deep reinforcement learning system for sorting waste at scale (15 citations), Kappler’s research consistently pushes the boundaries of practical, generalizable robotic intelligence.
Research Focus
Key Achievements
Top Papers
- 1Open-vocabulary Queryable Scene Representations for Real World Planning125 citations · 2023
- 2Real-Time Perception Meets Reactive Motion Generation107 citations · 2018
- 3BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learning89 citations · 2022
- 4Learning where to search using visual attention56 citations · 2016
- 5
- 6Riemannian Motion Policies49 citations · 2018
- 7Robot arm pose estimation by pixel-wise regression of joint angles36 citations · 2016
- 8Templates for pre-grasp sliding interactions29 citations · 2011
- 9
- 10