Surgical WAM: A World-Action Model for Data-Efficient Surgical Robot Learning
2026-08-11 • Robotics
RoboticsArtificial IntelligenceComputer Vision and Pattern Recognition
AI summaryⓘ
The authors address the challenge of teaching surgical robots to perform tasks when there are few examples labeled with exact robot actions. They propose Surgical WAM, a model that first learns how surgical scenes change over time using many unlabeled surgery videos and then fine-tunes its understanding with a small set of action-labeled data. This method helps the robot plan and adjust its movements in real time, improving success rates on simulated surgical tasks, especially those needing delicate contact and coordination between two hands. Their work suggests that learning from lots of unlabeled video can make training surgical robots more efficient.
Surgical robotAction-labeled demonstrationsEndoscopic videoWorld modelsClosed-loop controlCosmos PolicyVisual dynamicsReceding-horizon controllerBimanual coordinationVideo pretraining
Authors
Wenrui Bao, Tianyun Jiang, Zhiben Chen, Ser-Nam Lim, Peter D. Peng, Yuzhang Shang
Abstract
Learning reliable surgical manipulation policies is bottlenecked by the scarcity of action-labeled demonstrations: teleoperated surgical robot (e.g., dVRK) trajectories with synchronized kinematics are costly to collect, while surgical tasks demand precise contact handling, long-horizon reasoning, and bimanual coordination. Endoscopic video is comparatively inexpensive and abundant relative to synchronized video--kinematics trajectories, and a natural way to exploit it is to learn world models of surgical scenes. However, existing surgical world models use video primarily for simulation or policy evaluation, and rarely translate the learned dynamics into closed-loop control. This gap raises our central question: under a fixed budget of action-labeled demonstrations, does action-free video pretraining improve closed-loop surgical manipulation? To answer it, we introduce the Surgical World-Action Model (Surgical WAM), a unified generative model built on Cosmos Policy that jointly predicts future endoscopic observations and executable surgical robot action chunks. Surgical WAM first learns surgical visual dynamics from action-free video and is then fine-tuned on the fixed action-labeled budget; at deployment, it acts as a closed-loop, receding-horizon controller that executes a short prefix of each predicted action chunk and replans from the resulting observation. On a suite of four simulated surgical manipulation tasks, video pretraining improves the average success rate from 63.5% to 77.8%, including an absolute gain of 20 percentage points on PegTransfer, with the largest improvements on contact-rich and bimanual tasks. These results demonstrate that action-free video provides transferable visual dynamics priors for learning surgical robot control with limited action supervision, positioning data-efficient video pretraining as a practical path toward scaling up surgical robot learning.