🤖 Surgical Robotics AI Trainer

Demonstrating the synergy between robotic precision and AI safety. Control a virtual surgical arm with tremor filtration, or watch the AI perform autonomous tasks with perfect precision.

📚 Related Datasets & Resources

This demonstrator is inspired by real-world surgical robotics research. Explore these open datasets that enable AI-driven surgical tool detection, pose estimation, and action recognition:

  • SARAS-MESAD Dataset — Digital recordings from da Vinci Xi robotic system with 21 action classes (MESAD-Real) and 14 action classes (MESAD-Phantom) for prostatectomy procedures. Contains ~23K training images from real patient procedures and phantom training sessions.
  • SurgPose Dataset — ~120K surgical instrument instances (6 categories) with 7 semantic keypoints each, enabling 3D pose estimation and tracking for robotic surgical tools. Collected using UV-reactive paint for precise annotation.
  • Surgical Tool Datasets Repository — Comprehensive curated list of surgical tool datasets organized by task (tool classification, segmentation, action detection, skill assessment) including Cholec80, EndoVis, JIGSAWS, and more.

Mode Selection

Difficulty Level

Robotic Systems

AI Safety Monitor

Precision Score: 100%
Blood Loss: 0 ml
Progress:
Time Elapsed: 00:00
Tremor Level: Low
Safety Violations: 0
🏥 Controls:
Mouse Move Surgical Arm
W A S D Move Coagulator
Space Cauterize Bleeding
Enter Finish Task
System Active: Master-Slave Engaged