# LiDAR

LiDAR sensors provide high-precision 2D planar scans and 3D point cloud perception for autonomous mobile robots (AMRs), navigation, and obstacle avoidance.

## Validated Hardware

The following LiDAR sensors have been validated for use with the Robotics AI Suite:

| Sensor | Type | Interface | Output Topic(s) | ROS 2 Driver Package |
|---|---|---|---|---|
| **Velodyne Puck (VLP-16)** | 3D LiDAR (16-channel) | Ethernet (UDP) | `/velodyne_points` (`sensor_msgs/msg/PointCloud2`) | [`velodyne`](https://github.com/ros-drivers/velodyne) |
| **RoboPeak / Slamtec RPLIDAR** | 2D LiDAR (360° Planar) | USB (Serial UART) | `/scan` (`sensor_msgs/msg/LaserScan`) | [`sllidar_ros2`](https://github.com/Slamtec/sllidar_ros2) / `rplidar_ros` |
| **SICK nanoScan3 Pro** | 2D Safety LiDAR | Ethernet (COAP/UDP) | `/scan` (`sensor_msgs/msg/LaserScan`) | [`sick_safetyscanners2`](https://github.com/SICKAG/sick_safetyscanners2) |

---

## Integration Overview

LiDAR integration is robot and driver dependent. Complete driver setup and verify sensor data before integrating with navigation or perception stacks.

### 1. Driver Installation & Bringup

Install and configure the ROS 2 driver corresponding to your sensor interface:

::::{tab-set}
:::{tab-item} **Velodyne Puck (VLP-16)**
:sync: velodyne

The Velodyne Puck connects over Ethernet and streams raw UDP packets that are unpacked into 3D point clouds.

- **Driver:** [`velodyne`](https://github.com/ros-drivers/velodyne)
- **Install (binary):**
  ```bash
  sudo apt install ros-${ROS_DISTRO}-velodyne
  ```
- **Launch:**
  ```bash
  ros2 launch velodyne velodyne-all-nodes-VLP16-launch.py
  ```
- **Default Output:** `/velodyne_points` (`sensor_msgs/msg/PointCloud2`)

:::
:::{tab-item} **RoboPeak / Slamtec RPLIDAR**
:sync: rplidar

RoboPeak/Slamtec 2D scanners connect over USB serial (typically `/dev/ttyUSB0`) to publish 360-degree planar scan profiles.

- **Driver:** [`sllidar_ros2`](https://github.com/Slamtec/sllidar_ros2) (or `ros-${ROS_DISTRO}-rplidar-ros`)
- **Install (binary):**
  ```bash
  sudo apt install ros-${ROS_DISTRO}-rplidar-ros
  ```
  *(Or clone and build `sllidar_ros2` from source for newer A/S-series models.)*
- **Launch:**
  ```bash
  ros2 launch sllidar_ros2 sllidar_launch.py
  ```
- **Default Output:** `/scan` (`sensor_msgs/msg/LaserScan`)

:::
:::{tab-item} **SICK nanoScan3 Pro**
:sync: sick

The SICK nanoScan3 Pro is an industrial safety LiDAR communicating over Ethernet via the SICK safety scanner protocol.

- **Driver:** [`sick_safetyscanners2`](https://github.com/SICKAG/sick_safetyscanners2)
- **Install (binary):**
  ```bash
  sudo apt install ros-${ROS_DISTRO}-sick-safetyscanners2
  ```
- **Launch:**
  ```bash
  ros2 launch sick_safetyscanners2 sick_safetyscanners2_launch.py sensor_ip:=<SENSOR_IP>
  ```
- **Default Output:** `/scan` (`sensor_msgs/msg/LaserScan`)

:::
::::

### 2. Sensor Placement & Frame Transforms (TF)

Confirm that static transforms for your LiDAR optical/mounting frames (for example, `laser` or `velodyne`) are broadcast relative to `base_link`:

- Review the relevant [Hardware Blueprint](../../hardware_blueprints/index.md) (such as the [Clearpath Jackal AMR](../../hardware_blueprints/amr/clearpath-jackal.md)) for mounting geometries and transform definitions.
- Verify published frames and data rates:
  ```bash
  ros2 topic hz /scan
  ros2 run tf2_tools view_frames
  ```

---

## Applications & Pipelines

Once verified, LiDAR streams feed directly into Robotics AI Suite perception and navigation components:

- **[LiDAR & Visual Odometry Pipelines (LIO & LIVO)](../navigation/lio-livo-pipelines.md):** 6-DoF state estimation and SLAM using 3D LiDAR point clouds (FAST-LIO2, Point-LIO).
- **[3D Pointcloud Groundfloor Segmentation](reference_applications/pointcloud-groundfloor-segmentation.md):** Segment ground planes, ramps, and obstacle point clouds for traversability analysis.
- **[Nav2 Integration](../navigation/nav2-integration.md):** Feed 2D `/scan` and 3D voxel layers into standard costmaps and planners.