JetAuto: One Platform to Unlock SLAM, Vision, Voice, and Control

JetAuto is equipped with high-performance LiDAR and multi-modal sensors, transforming Simultaneous Localization and Mapping (SLAM)—the foundational technology for autonomous robot mobility—into an accessible learning experience. It supports both classic and cutting-edge algorithms such as Gmapping and Cartographer, enabling users to personally engage in the complete workflow: environmental exploration, map construction, path planning, and dynamic obstacle avoidance. What stands out even more is its industry-leading integration of “AI large models with SLAM.” By connecting to large language models like Qwen, JetAuto can understand natural language commands such as “Go to the zoo to see the animals.” It autonomously analyzes the semantic meaning of the environment and plans its route, elevating navigation from traditional coordinate-based movement to a new level of intent understanding. Behind this capability, its patented pendulum-suspended Mecanum wheel chassis ensures stability and precision on complex surfaces, providing a reliable hardware foundation for algorithm validation.

If LiDAR gives robots the ability to perceive spatial outlines, the 3D structured-light depth camera on JetAuto opens the door to understanding object details and three-dimensional relationships. It can acquire depth images and point cloud data in real time, enabling object recognition, distance measurement, volume analysis, and even 3D scene reconstruction. Combined with deep learning frameworks like YOLOv11, users can work on AI vision projects such as object detection and visual tracking. When expanded with a vision-enabled robotic arm, this visual system can be upgraded to “hand-eye coordination,” enabling autonomous grasping and handling based on 3D recognition. This seamlessly bridges visual perception with physical operation, offering a deep dive into the closed-loop logic of embodied intelligence.

The future of intelligent robots is inseparable from natural human-machine interaction. JetAuto’s integrated 6-channel microphone array not only supports high-precision sound source localization and noise reduction but is also deeply integrated with AI large models, achieving genuine voice-based interaction. Users can control the robot’s movement, navigate it to specific locations, or engage in Q&A dialogue—all through natural language commands. This is more than just an application of speech recognition technology; it reflects the “comprehension and reasoning” capabilities that large language models give to robots, transforming them from mere executors of coded commands into intelligent partners that can understand intent and even hold multi-turn conversations.

JetAuto’s capabilities do not stop at single-unit operation. It supports multi-robot communication and formation control, opening the door for learners to explore multi-robot systems. Whether it’s multi-robot cooperative navigation, formation maintenance (such as single-file line or triangular formation), or group control via a single joystick, these functions provide excellent experimental scenarios for studying distributed control and collaborative task planning. Through such hands-on practice, learners can intuitively understand the complex logic of communication, coordination, and obstacle avoidance in robot networks, gaining key experience for developing larger-scale automated systems in the future.

With JetAuto, Hiwonder is essentially delivering a complete “learning ecosystem.” It integrates all elements—from the Jetson Nano/Raspberry Pi main controller and high-precision sensors to fully open-source ROS 1/ROS 2 code and course materials. Learners can follow over 325 original lessons, starting from motor control and sensor drivers, and gradually advance to SLAM algorithm tuning, visual model training, voice interaction integration, and multi-robot cooperative algorithm development. This truly enables end-to-end skill mastery, from theory to real-world application. JetAuto proves that mastering cutting-edge robotics technology doesn’t necessarily require assembling individual components or setting up a complex lab—a well-designed, fully functional robot can be your most capable partner in exploring the intelligent world.