🛠 Advanced Usage — Hardware Upgrades

This section covers compatible hardware upgrades for your LIMO system, how to integrate them into your ROS 2 setup, the performance gains you can expect, and how to troubleshoot common upgrade issues.

---

🔌 Compatible Hardware Upgrades

Component	Upgrade Option	Benefit
Camera	Intel RealSense D435i, ZED 2i, Orbbec Astra+	Higher depth accuracy, wider FOV, better low-light performance
LiDAR	RPLIDAR A3, Hokuyo UST-10LX	Improved SLAM, faster map updates
Arm	MyCobot Pro 600, UFactory xArm 6	Higher payload, reach, and precision
Compute	NVIDIA Jetson Orin Nano / Xavier NX	Faster YOLO inference, improved multi-node performance
Networking	Wi-Fi 6 or Gigabit Ethernet module	Lower latency for remote monitoring
Power	Higher capacity Li-ion battery pack	Extended mission runtime

Pro Tip

If you upgrade your camera or LiDAR, always remap and recalibrate your environment before running autonomous missions.

---

🔄 Integration Procedures

Upgrading the Camera

1. Install Drivers

   • Intel RealSense:

     sudo apt install ros-foxy-realsense2-camera

   • ZED:
     Install SDK from stereolabs.com.

2. Update Launch Files

   • Replace orbbec_camera/dabai.launch.py with the correct camera node.
   • Update topics in object_detector.py:

     rgb_topic = '/camera/color/image_raw'
     depth_topic = '/camera/depth/image_raw'

3. Recalibrate in RViz:

   • Use tf2_ros static transform publisher to set correct camera → base_link offset.

Upgrading the LiDAR

1. Install ROS 2 Driver Package
   For RPLIDAR:

   sudo apt install ros-foxy-rplidar-ros

2. Update SLAM Launch
   Edit RTAB-Map or Cartographer launch to use /scan from new LiDAR.

3. Test:

   ros2 topic echo /scan

Upgrading Compute Unit

1. Install OS & ROS 2 on new board.
2. Copy your workspace:

   rsync -avz krish_ws/ user@new_board:/home/user/

3. Rebuild:

   colcon build --symlink-install

4. Test YOLO inference speed with:

   yolo detect predict model=best.pt source=sample.jpg

---

🚀 Performance Improvements

• Camera Upgrade → Better depth accuracy = more precise /target_pose
• LiDAR Upgrade → Faster loop closures and cleaner maps
• Compute Upgrade → Lower detection latency, higher navigation update rate
• Battery Upgrade → 30–50% longer missions
• Network Upgrade → More stable RQT/remote teleop feeds

---

🛠 Upgrade Troubleshooting

Symptom	Possible Cause	Solution
Camera node crashes	Driver mismatch	Install correct ROS 2 driver for your camera
Depth image blank	Wrong topic names	Update object_detector.py to new depth topic
Nav2 fails after LiDAR upgrade	Map resolution mismatch	Remap environment with new sensor
YOLO slower after upgrade	Wrong CUDA setup	Reinstall NVIDIA drivers & torch with GPU support
RQT feed freezes	Network bottleneck	Switch to wired Gigabit Ethernet for testing

Important

After any hardware change, clear your old maps and re-run full mapping to ensure accurate localization.

---

Next: 🛡 Troubleshooting & FAQ