Read the paper here:


  • This work was submitted as a “blue-sky” workshop paper, where the goal is to outline some big idea or take some kind of position in terms of a research direction or agenda.
  • Here, I talk about the benefits of object-level planning, where a robot needs to start with some schematic plan that then is broken down into task- and motion-level planning (as in TAMP).


D. Paulius (2022). “Object-Level Planning and Abstraction”. In: Conference on Robot Learning (CoRL) 2022 Workshop on Learning, Perception, and Abstraction for Long-Horizon Planning, .