Papers
7
Total Citations
125
H-Index
4
About
Aidan Curtis is a rising star in artificial intelligence and robotics, whose research lies at the intersection of task and motion planning, robot manipulation, and machine learning. His work is distinguished by a focus on enabling robots to operate intelligently in complex, real-world environments with limited information. Curtis’s most impactful contribution, with 62 citations, is his pioneering use of Graph Neural Networks to learn object importance for planning in large-scale problems, dramatically reducing computational complexity by identifying only the objects essential for finding a plan. He has further advanced the field by developing systems that combine general-purpose task-and-motion planning with learned affordances for long-horizon manipulation of unknown objects (35 citations), and by creating methods for discovering state and action abstractions that enable generalized planning across entire problem domains (15 citations). His recent work tackles critical challenges in visibility-aware navigation among movable obstacles and task-directed exploration in continuous POMDPs for articulated object manipulation. Curtis’s integrated approach—blending classical planning with modern learning techniques—is shaping the next generation of autonomous robotic systems capable of performing complex, contact-rich tasks like assembly with minimal human intervention.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Visibility-Aware Navigation Among Movable Obstacles5 citations · 2023
- 5
- 6
- 7