本页目录

XGO-lite3/mini2S

XGO-lite3/mini2S: World’s First Raspberry Pi Robotic Dog with An Arm

XGO-lite3 and XGO-mini2S are Raspberry Pi CM5 quadruped robots with integrated robotic arms, for education, AI interaction and robotics development.

In 2023 KickStarter, XGO 2, World’s First Raspberry Pi Robotic Dog with An Arm

1.Quadruped Platform (lite3 & mini2S)

Both models have 12 leg joints for locomotion and body posture control. XGO-lite3 has a 3-DOF arm (15 actuators in total); XGO-mini2S has a 4-DOF arm (16 actuators in total). They share the CM5 and ESP32 control architecture, with different memory, actuators and power systems.

Feature

XGO-lite3

XGO-mini2S

Leg joints / arm

12 / 3 DOF

12 / 4 DOF

Computer

Raspberry Pi CM5, 2 GB

Raspberry Pi CM5, 4 GB

Motion controller

ESP32

ESP32

Servo torque

2.3 kg·cm

4.5 kg·cm

Servo feedback

Potentiometer

Magnetic encoder

Servo range

0–300°

0–360°

Structure

1 mm aluminum alloy

1.5 mm aluminum alloy

Legs and feet

ABS

Silicone + ABS

Weight

610 g

915 g

Default standing dimensions

250 × 145 × 170 mm

270 × 150 × 180 mm

Battery

18650, 2S, 2500 mAh

18650, 2S, 3800 mAh

Charging

5 V USB Type-C

Supplied 8.4 V, 4 A DC charger

Typical runtime

Approximately 2 hours

Approximately 2 hours

Table.1 Selection Matrix: Education vs. Research

⚠️ CRITICAL POWER WARNING

XGO-lite3 is designed for convenience: Charge via standard USB Type-C (5V).

XGO-mini2S requires performance power: Charge via the included 8.4V 4A DC Charger.

Fig.1 XGO-lite3

Fig.2 XGO-mini2S

2. Detailed Specifications

2.1 XGO-lite3: The Agile Classroom Companion

A lightweight, fully assembled platform ready for Python and Blockly education.

  • Processor: Raspberry Pi CM5, 2 GB.

  • Sensors: 6-Axis IMU, Joint Position Sensors.

  • Vision/Audio: 2.0" IPS Screen, 5MP Camera (OV5647), Dual MEMS Mic, 8Ω2W Speaker.

  • Key Features:

    Quiet Operation: Optimized for classroom environments.

  1. Move & Record: Teach the robot movements by hand-guiding the joints.

  • Included Kit: Protective Case, Type-C Hub, HDMI Cable, Color Blocks.

2.2 XGO-mini2S: The High-Performance Lab Platform

Engineered for advanced algorithm verification and secondary development.

  • Processor: Raspberry Pi CM5, 4 GB.

  • Advanced Actuation: Uses Hollow Cup Motors (Coreless) for rapid response and Magnetic Encoders for 0.01° accuracy.

  • Vision/Audio: 2.0" IPS Screen, 5MP Camera (OV5647), Dual MEMS Mic, 8Ω2W Speaker.

  • Key Features:

  1. Omnidirectional Motion: Decoupled kinematics allowing strafing and turning simultaneously.

  1. High Torque: Capable of dynamic gaits and carrying heavier sensor payloads.

  • Included Kit: Aluminum Case, 8.4V 4A Fast Charger, Type-C Hub, Color Blocks.

2.3 XGO-mini2S with LiDAR

Fig.3 XGO-mini2S with LiDAR

Optional LiDAR configuration: EAI YDLIDAR T-mini PLUS. The specifications below describe that sensor; the display range in an application is a separate setting.

Parameter

Value

Notes

Ranging Radius

0.05 - 12m

@80% reflectivity; 0.05 - 4m @10% reflectivity

Ranging Accuracy

±20mm

Full range systematic error

Scanning Frequency

6 - 12Hz

Software adjustable, default 6Hz

Ranging Frequency

4000Hz

High sampling rate

Scanning Angle

360°

Omni-directional scanning

Ambient Light Resistance

60000 Lux

Strong anti-interference ability

Additional Key Specifications

  • Angle Resolution: 0.54°

  • Laser Safety: FDA Class I (Eye safe)

  • Supply Voltage: 4.8V - 5.2V (Typical 5.0V)

  • Communication Interface: UART / GH1.25-4P

  • Dimensions: 38.6mm × 38.6mm × 33.9mm

  • Weight: Approx. 45g

3. Core Technologies

Despite their mechanical differences, both robots share Luwu Intelligence's core control stack:

  • IMU Self-Stabilization: Closed-loop attitude control ensures the chassis remains level even when the surface tilts or moves.

  • 6-DOF Posture Control: The feet remain fixed while the body adjusts Pitch, Roll, Yaw, Heave, Surge, and Sway.

  • Kinematics Simulation: Fully compatible with Rviz / Gazebo for simulation-to-reality (Sim2Real) workflows.

  • AI Ecosystem: Native support for Machine Vision (Face/Gesture), Speech Recognition, and Secondary Development via Serial Protocol.

Model generations

The K210-based XGO-lite / XGO-mini and CM4-based XGO-lite2 / XGO-mini2 are discontinued models. Their historical names and instructions should not be confused with the current lite3 / mini2S specifications. Select system images by the exact product and hardware revision; compatible CM4 hardware may also run a newer Luwu-OS release.