LanderPi Hardware: 5 Genius Design Choices You Might Have Missed

LanderPi combines the power of Raspberry Pi 5 as the decision-making brain with the precision of STM32 as the control unit, creating an efficient “perception-decision-control” loop. The Raspberry Pi 5 handles large-scale data from LiDAR, 3D depth camera, and AI models, performing tasks like SLAM mapping, path planning, and visual recognition. Meanwhile, STM32 ensures real-time motor and servo control, providing precise execution.

This dual-brain setup not only complements the system's performance but also ensures stability and scalability. With this collaboration, LanderPi can tackle complex AI tasks while maintaining millisecond-level response for seamless movement and control, laying a solid foundation for functions like chassis movement and robot arm operations.

With its control architecture in place, LanderPi takes its versatility a step further by offering three chassis configurations, making it adaptable to a wide range of applications. It supports Mecanum wheels, Ackermann steering, and Tank chassis, each suited to a unique application. Mecanum wheels allow for 360° omnidirectional movement, perfect for navigating tight spaces.

The Ackermann steering system mimics the turning mechanism of a real car, making it ideal for autonomous driving research. The Tank chassis, equipped with Hiwonder's high-elastic torsion spring suspension, effortlessly handles rugged terrains, offering full-terrain mobility. This multi-chassis flexibility goes beyond just mechanical compatibility. It seamlessly integrates with ROS2 motion control algorithms, offering a comprehensive platform that allows for hands-on learning and practice across everything from flat surfaces to challenging terrains.

Flexibility in movement is one aspect, but true intelligence lies in the ability to interact with precision. In grasping tasks, seeing does not necessarily mean grabbing. LanderPi takes this to the next level by combining a 3D depth camera with a 6DOF robot arm. Through advanced inverse kinematics, this integration enables authentic hand-eye coordination. This is not merely a combination of a camera and a robot arm —it is a complete system where vision and action are perfectly synchronized. The robot can pinpoint an object’s exact 3D position and instantly calculate the necessary movements to grasp and manipulate it. Whether it's target tracking, shape recognition, or using VSLAM for visual navigation, LanderPi delivers capabilities far beyond basic educational applications, demonstrating true robotic intelligence.

Precision vision and motion planning are crucial, but the final task of grasping falls to the end-effector—the robotic gripper. Despite its compact size, LanderPi’s gripper features a biomimetic rail mechanism developed by Hiwonder. Its parallel rail design, combined with a metal linkage structure, ensures stable horizontal gripping and a wide opening range. The EVA anti-slip pads on the gripper tips provide a secure grip, even on delicate items. These seemingly small design details are the result of extensive testing and fine-tuning. The rail gripper can handle objects up to 500g, yet remains precise enough to grasp delicate and smaller items. This meticulous attention to design makes the gripper a highly reliable tool, ensuring flawless precision in those crucial final moments of robotic interaction.

After mastering precise movement and handling, the next challenge is making human-robot interaction feel natural and intuitive. LanderPi solves this with its integrated voice interaction system powered by advanced AI models. With an AI voice module, the robot can connect to large models via OpenAI, giving it the ability to perceive its environment, plan tasks, and act autonomously. You can give the robot simple voice commands like, “Hiwonder, bring the red package from the drop-off point to my house.” It not only recognizes the red package, but also understands the intent behind the action, such as bring it home, and autonomously breaks the task into steps: move to the drop-off point, pick up the red package, and return home. This is a major leap toward more advanced, embodied AI applications.

The five key innovations in LanderPi’s design follow a clear, unified philosophy: starting with stable, efficient control, progressing to versatile movement capabilities, followed by precise perception and manipulation, and culminating in smooth, intuitive human-robot interaction. These features are not isolated, but rather an integrated system where each element supports and enhances the overall performance.