The simulation lacked human obstacles for LiDAR to detect, and the camera viewport had no object-detection visualization.

Humans

• 5 NPC stick figures spawn alongside blocks when Chassis is added; wander with random direction changes and wall bouncing
• getLidarDetection() extended to scan humans (circle collision via pointToSegmentDist) after block scan; returns { ...obj, type: 'block' | 'human' }
• Detection panel shows "● HUMAN DETECTED" (orange) vs "● OBJECT DETECTED" (green)

Camera object detection & tracing

• drawCameraView() refactored with a shared project() helper for perspective projection
• Blocks get cyan bounding boxes + shape label (Square / Circle / Triangle)
• Humans render as grey silhouettes with orange bounding boxes + "Human" label
• Box size scales with distance via the existing perspective projection

// LiDAR now checks humans after blocks
for (const h of humans) {
  if (pointToSegmentDist(h.x, h.y, lidarX, lidarY, beamEndX, beamEndY) < h.size / 2) {
    return { ...h, type: 'human' };
  }
}

Screenshots

Humans wandering on the main canvas with the robot:

LiDAR + Camera active — bounding box with shape label visible in camera viewport:

🔒 GitHub Advanced Security automatically protects Copilot coding agent pull requests. You can protect all pull requests by enabling Advanced Security for your repositories. Learn more about Advanced Security.