Overview

Robot-assisted surgery (RAS) is now increasingly deployed and delivers clear gains in precision, stability, and ergonomics, in comparison to conventional approaches. Yet these advancements also reveal several bottlenecks, such as sustaining optimal views, exposure, and coordination, still demanding intensive micromanagement by the surgeon amid patient and workflow variability, work that typically falls to an expert assistant. To reduce reliance on a dedicated assistant and enable a reliable single-surgeon workflow, we focus on “Intelligent Robotic Assistants” that fuse Embodied-AI perception, surgeon-intention recognition, and verifiable shared/autonomous control to offload camera steering, atraumatic tissue exposure, and context-aware guidance during surgical task execution.

Specifically, recent advances in embodied intelligence, spanning advanced control frameworks, learning-based perception and policy optimization, and generative models for planning, offer a concrete path to address these challenges. By convening academic researchers, industry, and clinicians, this workshop will catalyze cross-disciplinary progress toward next-generation “Robotic Assistants” across diverse surgeries and interventional procedures. In parallel, we will examine opportunities, challenges, and ethical/regulatory considerations, and define evaluation protocols and translational pathways to convert these technical advances into clinically viable, trustworthy solutions.