NavWareSet Tutorials

📚 NavWareSet Tutorials is a hands-on workspace showing how to open, visualize, and work with data from the NavWareSet dataset.

This repository demonstrates:

• How to plot robot & participant trajectories from CSV.
• How to visualize annotated 3D point clouds and occupancy maps.
• How to replay LIDAR data from .bag files with proper static transforms.
• How to inspect LIDAR data using RViz.

---

📂 Repository Structure

.
├── 13_annotated/            # Example annotation files
├── 13_poses/                # Example pose CSVs
├── 13_grs.bag               # Example ROS bag (GRS sensors)
├── 13_robot.bag             # Example ROS bag (robot sensors)
├── csv_scene_viewer.py      # Python: 2D pose & occupancy plot
├── pcd_annotated_scene_viewer.py  # Python: 3D PCD + annotations
├── launch_lidar_data.launch # ROS launch file for bag playback & RViz
├── my_rviz.rviz             # Example RViz config
├── requirements.txt         # Python dependencies
├── README.md                # 📖 You are here!

---

Disclaimer: The ROS bags in this repo are 1 second slices of the originals. The originals are dozens of GB in size and because of that can not be uploaded to GitHub.

⚙️ 1️⃣ Install Python Dependencies

Create a virtual environment (recommended):

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

---

▶️ 2️⃣ Run the Python Scripts

✅ A. csv_scene_viewer.py — Plot 2D Poses & Occupancy

What it does:
Plots the robot’s trajectory, participant positions, and the occupancy map over time using Matplotlib.

How to run:

source .venv/bin/activate
python csv_scene_viewer.py

This will open an animated Matplotlib window showing robot movement and static occupancy points.

---

✅ B. pcd_annotated_scene_viewer.py — View 3D Point Clouds + Annotations

What it does:
Visualizes time-synchronized 3D point clouds, annotated bounding boxes, robot frames, and occupancy grids using Open3D.

How to run:

source .venv/bin/activate
python pcd_annotated_scene_viewer.py

Controls:

• → / ← : Step forward/backward through frames.
• W/A/S/D, Q/E, ↑/↓ : Adjust occupancy grid position and yaw.
• P : Save current point cloud.
• O : Save occupancy grid XY points.
• Space : Exit viewer.

---

📡 3️⃣ Open the ROS Bags

✅ A. 13_grs.bag — Play LIDAR Data with Transforms

How to run:

roslaunch launch_lidar_data.launch

What this does:

• Plays the 13_grs.bag at 1x speed.
• Publishes required static transforms (map → rslidar → camera_color → camera_color_optical_frame).
• Opens RViz with my_rviz.rviz configuration.

Tip:
If you want to use your own .bag or .rviz config, edit launch_lidar_data.launch or pass them as arguments.

---

✅ B. 13_robot.bag — Play Robot Poses (Manually)

How to run:

This bag can be played separately:

rosbag play --clock 13_robot.bag

Then open RViz manually:

rviz -d my_rviz.rviz

Note: Make sure use_sim_time is true if playing bags with timestamps.

---

📎 4️⃣ Dataset Source

To download or explore more scenes, visit:
👉 NavWareSet Official Website

---

✅ 5️⃣ Best Practices

• Always activate your Python virtual environment:

  source .venv/bin/activate

• Keep ROS bags out of Git if they are large — share them via dataset download or separate link.
• Use relative paths in your .launch files for portability.

---

📖 6️⃣ Citation

If you use NavWareSet in your research, please cite the authors as indicated on the official website.

---

🚀 Happy exploring!