RViz is ROS's 3D visualization tool — it shows what a robot sees and thinks, from lidar scans and camera feeds to planned paths and coordinate frames, making the invisible state of a robot visible for debugging.
RViz is a 3D window that shows what a robot is sensing and planning — its lidar points, camera image, the map, and where it intends to go. It turns a robot's invisible internal data into something you can see and debug.
A robot's inner life is invisible — streams of numbers for lidar points, transforms, and plans. RViz turns all of it into a 3D picture you can actually look at, and it's the first tool most engineers open when something's wrong.
What it does
RViz subscribes to topics and renders them in 3D: the lidar's point cloud, the camera image, the robot's own model (from its URDF), the tf coordinate frames, the occupancy map, the planned path, detected objects. You assemble a view by adding "displays," each tied to a topic. Whatever the robot is publishing, RViz can show.
Topics in, 3D picture out
RViz is a window onto the robot's data — lidar, frames, paths, maps — overlaid in one 3D scene so mismatches jump out visually.
Why it's essential for debugging
Most robot bugs are spatial, and RViz makes them obvious:
A transform is wrong? The lidar points appear rotated or floating off the robot.
The map is misaligned? You see it drift against the scan.
The planner picked a weird path? It's drawn right there.
A sensor is dead? Its display goes empty.
Pair RViz with rosbag and you can replay and watch a past run frame by frame — the standard way to diagnose a failure after the fact.
RViz vs Gazebo — a key distinction
Beginners often confuse them. RViz visualizes what the robot believes (its published data); Gazebo simulates physics to generate data as if from a real world. RViz doesn't move the robot or model gravity — it's a viewer, not a simulator. You often run both: Gazebo produces sensor data, RViz shows what the robot makes of it.
Why it matters
RViz is the eyes of a ROS developer — the tool that makes a robot's perception, transforms, and plans visible. Fluency with it is one of the most practical, everyday skills in robotics development.
