MSI’s MS-98L3 3.5" SBC Optimized MSI AMR-AI-Base-Robot, Taiwan

Background

MSI sought to develop a high-performance AMR-AI-Base-Robot for factory automation. The robot was request to be autonomous navigation, real-time map creation, and obstacle avoidance in narrow and cluttered spaces. Given the complexity of these tasks, AMR required a robust hardware platform capable of handling multiple sensors, real-time data processing, and integration with external devices.

Solution: MS-98L3 3.5” SBC

To meet the requirements, the MS-98L3 board was selected due to its powerful CPU, flexible connectivity options, and expandability. This allowed the integration of several essential components for the AMR, including laser sensors (LiDAR), 3D cameras, motor control, and Wi-Fi connectivity.

Key Hardware Configuration:

• Board Model: MS-98L3
• Connectivity:
  • 2 LAN ports connecting to LiDAR laser sensors for obstacle detection and environment scanning.
  • 1 LAN port for connecting a Wi-Fi Access Point, supporting point-to-point communication.
  • 1 reserved LAN port for future external device integration, such as a robotic arm control box.
  • 3 USB 3.0 ports for connecting RealSense 3D cameras, enabling detailed environment mapping and navigation.
  • 2 COM (RS485) interfaces connected to the motor control box for precise movement control.
• Performance:
  • 8-core CPU: The system typically utilizes 2 cores during normal operations but scales up to 4 cores during intensive tasks like mapping.
  • Expandable RAM: While the specific size was not mentioned, expandability was a key reason for choosing the MS-98L3 SBC, ensuring future-proofing for memory-intensive applications.

Challenges

• Real-Time Map Creation: The AMR needed to generate detailed maps of its environment as it moved, ensuring accurate navigation. The LiDAR and 3D cameras continuously scanned the surroundings, feeding real-time data to the system for map creation and obstacle detection.
• Obstacle Avoidance and Collision Prevention: To ensure safety, the robot was equipped with laser sensors and 3D cameras. When an obstacle, such as a person, was detected within a 1.2-meter range, the system would slow down to prevent collisions.
• Space Constraints: The robot was designed to move through narrow spaces, making compact hardware crucial. Additionally, the height of the robot prevented it from navigating under shelves, a design compromise based on the target environment (offices, factories, warehouses and indoor settings).

Results

The MS-98L3 delivered the processing power and flexibility needed for this AMR application. Key outcomes included:

• Seamless Navigation: The AMR successfully navigated narrow corridors, avoided obstacles, and generated accurate environmental maps in real time, making it well-suited for office environments.
• Improved Efficiency: The robot's ability to perform tasks autonomously reduced the need for human intervention, optimizing workflow in the target environments.
• Durability and Reliability: The addition of an isolation design to the circuit board eliminated power-related issues with the COM interfaces, ensuring long-term reliability.
• Scalability: The system’s modular design allowed for future expansions, including the potential integration of additional peripherals like robotic arms.

Conclusion

The MS-98L3 board provided the ideal foundation for the AMR project, addressing key challenges like space constraints, real-time processing, and durability. Its flexibility and expandability make it an excellent choice for applications requiring robust hardware solutions in challenging environments.

This case highlights the importance of selecting the right hardware for mission-critical applications in automation and robotics. With the MS-98L3 board, MSI AMR was able to create a highly efficient and reliable autonomous mobile robot, improving overall operations in indoor environments.