We are excited to share that our ecosystem partner, UniRobot, has launched the brand new RAI-P4 Embodied Robot Task Planning Comprehensive Training Teaching Platform powered by D-Robotics’ RDK X5. Designed for AI and robotics education in higher institutions, this platform offers an open-source, out-of-the-box solution that bridges the gap between theory and real-world application. It empowers students to move beyond fragmented learning toward integrated project development.

Full-Stack Integration, Redefining Robotics Teaching

Centered around a 4-DOF robotic arm and built on the RDK X5 developer kit, the RAI-P4 leverages 10 TOPS of edge computing power to efficiently run AI tasks. It integrates large language models including Qwen and DeepSeek, and is even equipped with Qwen’s upgraded visual language model (VLM), enabling zero-shot recognition of new objects without pre-training which drastically lowering the barrier to visual perception.The platform also supports multiple AI-driven functions: voice interaction, semantic understanding, visual tracking, smart grasping, motion control, fruit and vegetable recognition, and more. It fully covers the entire embodied AI workflow from task planning and perception to execution. And this is ideal for hands-on educational use.

Fully Open-Sourced Core Functions for Deep Technical Learning

UniRobot is releasing open-source code for all core capabilities of the RAI-P4, offering full visibility into key technologies:

• Vision: Complete project code for face detection and object detection based on YOLOv8,as well as recognition of all things based on VLM
• Voice: End-to-end code for ASR, LLM, and TTS pipelines based on large language models
• Motion: Open-source underlying driver code for the robotic arm for deeper motion control insight

This allows students to transition from using technology to understanding it, building real development skills in intelligent robotics.

Out-of-the-Box Learning in a Compact 60x60 cm Setup

The RAI-P4 requires only a 60x60 cm desktop space for full operation. It comes pre-integrated with an AI motherboard, multiple sensors, a display module, and wireless keyboard/mouse. The system ships with Ubuntu, ROS 2, Python, and VScode pre-installed, eliminating complex setup. And students and educators can focus on learning from day one.

80+ Hands-On Lessons Tailored for Multi-Disciplinary Use

The platform includes 8 course modules and over 80 class hours of experimental content, covering kinematics control, YOLO visual deployment, LLM-based dialogue, ROS 2 development, and more. Each lesson comes with open-source code, 3D mechanical drawings, and is designed for seamless integration into programs such as AI, Robotics, Automation, and Computer Science.UniRobot specializes in bringing cutting-edge embodied AI into education through open curricula and community building. Its mission is to make robotics technology accessible to everyone through progressive learning, and this aligns closely with the vision of D-Robotics. For partnership inquiries or purchasing information, please contact UniRobot at uni_robot@outlook.com.

The launch of the RAI-P4 platform marks another milestone in our shared commitment to innovation in educational robotics and embodies our effort to empower universities in cultivating the next generation of embodied AI engineers.