News

Our lab has two papers that were accepted to the International Conference on Robotics and Automation (ICRA) 2026 conference! These papers are:

• Hybrid Diffusion Policies with Projective Geometric Algebra for Efficient Robot Manipulation Learning (authors: Xiatao Sun, Yuxuan Wang, Shuo Yang, Yinxing Chen, Daniel Rakita)

• Subsecond 3D Mesh Generation for Robot Manipulation (authors: Qian Wang, Omar Abdellall, Tony Gao, Xiatao Sun, Daniel Rakita)

Our paper titled “Deep learning approach for critical exposure during division of the inferior mesenteric artery in colorectal surgery” has been accepted for publication in the Journal of Robotic Surgery.

You can read the full article here: https://link.springer.com/article/10.1007/s11701-025-03121-7

Congratulations to the entire team!

Our lab has two papers that were accepted to the International Conference on Intelligent Robots and Systems (IROS) 2025 conference! These papers are:

• Towards Zero-Knowledge Task Planning via a Language-based Approach (authors: Liam Merz Hoffmeister, Brian Scassellati, Daniel Rakita)

• ad-trait: A Fast and Flexible Automatic Differentiation Library in Rust (Chen Liang, Peter Wang, Andy Xu, Daniel Rakita)

Roshan Klein-Seetharaman was awarded a National Science Foundation (NSF) Graduate Research Fellowship Award! Congratulations, Roshan!

Our lab has two papers that were accepted to the Robotics Science and Systems (RSS) 2025 conference! These papers are:

• Dynamic Rank Adjustment in Diffusion Policies for Efficient and Flexible Training (authors: Xiatao Sun, Shuo Yang, Yinxing (Jack) Chen, Franics Fan, Yiyan (Edgar) Liang, Daniel Rakita)

• Coherance-based Approximate Derivatives via Web of Affine Spaces Optimization (authors: Daniel Rakita, Chen Liang*, Qian (Peter) Wang*, where * means equal contribution)

We will share paper websites, pre-print articles, and videos soon!

Our work has been accepted for presentation at the Korean Society of Endo-Laproscopic & Robotic Surgery (KSERS) Symposium in April 2025. This work was done in collaboration with Yale New Haven Health.

Title: AI-Powered Semantic Segmentation Model for Enhanced Ureteral Mapping and Real-Time Instrument Feedback in Robotic Surgery

Authors: Jay Bader, Xiatao Sun, Alexis Ramirez-Hardy, Tripp Rosenfelt, Daniel Rakita, Amit Khanna

Our work has been accepted for presentation at the American Society of Colon and Rectal Surgeons (ASCRS) meeting in San Diego in May 2025. This work was done in collaboration with Yale New Haven Health.

Title: AI-Technology Innovations in Robotic Colorectal Surgery – Detecting Optimal Angles for Critical Exposure and Division of the Inferior Mesenteric Artery

Authors: Jay Bader, Xiatao Sun, Alexis Ramirez-Hardy, Tripp Rosenfelt, Daniel Rakita, Amit Khanna

Our paper Sequential Discrete Action Selectionvia Blocking Conditions and Resolutions (authors Liam Merz Hoffmeister, Brian Scassellati, and Daniel Rakita) will be presented at IROS 2024.

Our lab was highlighted in the 2024 edition of Yale Engineering. Explore the full magazine here for in-depth coverage of the exciting developments across Yale Engineering.

Professors Daniel Rakita and Brian Scassellati were awarded a grant for their proposal “Robot Manipulation in Densely Cluttered Environments”. This work will focus on creating a generalizable framework that enables robots to work in cluttered environments, such as for homecare, healthcare, agriculture, or disaster relief scenarios. Areas of investigation will include dynamically adjusting viewpoints to locate occluded objects, computing long action sequences to clear obstacles, and crafting motion strategies that make strategic contacts with surrounding items to gain better access to parts of the workspace. The $1.2M grant is funded by the Office of Naval Research.

News article on our paper An Analysis of Unified Manipulation with Robot Arms and Dexterous Hands via Optimization-based Motion Synthesis that will be featured at ICRA 2023 in London.

Our paper, Lively: Enabling Multimodal, Lifelike, and Extensible Real-time Robot Motion, won a Best Paper Award at HRI 2023! Check out documentation and code for our proposed method here.