This week you can find me in-person at #IROS2023, where I'll be co-organizing the workshop on “Task and Motion Planning: from Theory to Practice” (w/ Neil Dantam, Federico Pecora, and @LydiaKavraki)
Check out our phenomenal line-up of speakers here:
dyalab.mines.edu/2023/iros-w…
I am going on the academic job market! 🧑🎓
My research focuses on developing robot cognition 🤖
Among things, I contributed AI techniques to improve task & motion planning, decision making, and experience transfer.
If I might be a good fit for your department, please reach out!
@kingstonzak, @KavrakiLab and I have new work on massively accelerating robot motion planning (planning for the @FRANKAEMIKA Panda in 40 μs - about 500x faster than MoveIt!) on ordinary CPUs/SoCs: arxiv.org/abs/2309.14545
At #IROS2023? Come check out this forward-looking work from @DrRobotPhD and @kingstonzak interactions between robot motion planning and privacy! Today at 14:42 in the afternoon "Motion and Path Planning" session.
We had a blast attending #csbw2023 today at #ACMBCB2023@acm_bcb! Many great talks on structural bioinformatics in the post-AlphaFold2 era.
Our very own Felix Quintana presented his work on using Large Language Models and protein structure to predict pathogenicity. Check it out!
Rice CS team helps robots rearrange objects without gripping them in a new paper at #ICRA2023. The work is a collaboration between faculty members Kaiyu Hang & @LydiaKavraki, led by CS PhD student Kejia Ren with CS PhD student Podshara Chanrungmaneekul. bit.ly/3Jtuabq
Welcome 2023-24 MolSSI Software Fellow, Anja Conev @ConevAnja, a member of the @KavrakiLab at Rice U., who will be "developing a DINC-Ensemble toolkit for accelerated ensemble docking of large flexible ligands" during her year-long fellowship with us. Congratulations Anja! 👏😀
Rice CS' @KavrakiLab presented 4 #ICRA23#robotics & automation papers. “It's exciting to see our team members pushing the envelope...Few robotics & AI labs share this kind of access to top researchers with such different areas of expertise,” said Kavraki. bit.ly/3rQD78G
Choosing optimal placements and grasps can make a difference in cluttered environments for TAMP! If you are at #ICRA2023 in London, check out our poster presentation: Tue 05/30@3:00pm (Pod 40).
Amazing collaboration @KavrakiLab, @DrRobotPhD, @Tianyang_Pan, @kingstonzak
Yiyuan solves hard-to-model object rearrangement problems from only a start and goal configuration - no PDDL for actions or discretized state.
The planner embeds a physics engine and uses it to dynamically discover and search over stable intermediate arrangements.
This lets it automatically find valid pick-and-place actions to rearrange the objects - maintaining stability even on surfaces with challenging obstacle geometries, like the bumpy table shown here.
If you're in London, check out Yiyuan's poster at the 15:00 Weds session!
Excited to attend #ICRA2023 this week! 🤖
I will be presenting my ongoing work on automatic skill learning/transfer for life-long task and motion planning (w/ @KavrakiLab & @vardi).
Catch me on Wednesday morning's poster session and at the Compositional Robotics workshop Friday.
I'm going to be presenting this work (which is also published in RA-L) at #IROS2022 on Tuesday, Oct. 25!
Come by or get in touch if you're interested, and check out the project page: robotic-esp.com/papers/thoma…
Really excited about our work with my collaborators Michael Welle and Martina Lippi in IROS2022 on learning state representations for Visual Task Planning in the presence of variations that are task irrelevant.
Paper: arxiv.org/abs/2109.06737
Website: state-representation.github.…
Kavraki Lab's @kingstonzak was a finalist for the "Best Paper Award for Industrial Robotics with Real-World Applications" at #IROS2022! Congratulations to Zak!