Overview
DOGZILLA-Lite is a 15-DOF desktop AI bionic quadruped robot dog with an integrated robotic arm and end gripper for grasping and handling tasks. It uses an all-metal (aluminum alloy) body and is equipped with an integrated Raspberry Pi CM5 module for AI visual interaction, including face detection, object recognition, and multimodal large-model interaction (text, voice, and vision). A front 2.0-inch IPS display supports real-time interaction display and 35 dynamic expressions.
For product selection or technical questions, contact support at https://rcdrone.top/ or email support@rcdrone.top.
Key Features
- 15-DOF bionic design: 12DOF quadruped + 3DOF robotic arm (including end gripper) for flexible grasping and precise control.
- Posture stability and motion control: 15 high-precision 2.3KG·CM bus servos, 6-axis IMU attitude sensor, inverse kinematics algorithms; supports omnidirectional movement, 6D posture control, dynamic balance, and multiple movement gaits.
- Multimodal large-model interaction: large language model, voice large model (voice-to-text / text-to-voice interaction via microphone and speaker), and visual large model (image understanding and analysis with feedback).
- Integrated CM5 module interaction hardware: 2.0-inch IPS color display, four programmable buttons, 5MP HD camera, microphone, and speaker.
- Pre-installed GUI program: more than 40 functional functions for fast hands-on experience.
- Dual APP control: Bluetooth APP for movement/arm control; WiFi APP for real-time image transmission (FPV-style control experience).
- AI visual recognition examples: 3D object recognition, face detection, license plate recognition, target detection, emotion recognition, motion detection count, palm control, human skeleton recognition, face tracking, color following, QR code motion control, gesture following, gesture control (1/2/3/4/5/6/Good/OK/fist), and “brush” (finger drawing).
- Robotic arm related functions: APP remote control handling; visual recognition and handling; line tracking with obstacle removal using the arm; voice-command handling control via large language model.
- Built-in bionic actions: built-in 20 bionic action groups.
Specifications
| Product | DOGZILLA-Lite |
| Degrees of Freedom (DOF) | 15DOF |
| Legs / Arm DOF | 12DOF quadruped + 3DOF robotic arm (including end gripper) |
| Body material | Aluminum alloy / all-metal body |
| Main control | Raspberry Pi CM5 module (integrated) |
| Servo specification | 15 bus servos, 2.3KG·CM |
| IMU | 6-axis IMU sensor |
| Display | 2.0-inch IPS color display; 35 dynamic expressions; 320 x 240 pixel full color |
| Buttons | Four programmable buttons |
| Camera | 5MP HD camera; 5MP OV5647 camera |
| Microphone / speaker | Dual MEMS microphone; cavity speaker |
| Battery | 7.4V 2500mAh battery pack (2500mAh lithium battery) |
| Battery working time | 2.2 hours |
| Dimensions (power on) | 240.5*142.9*168.5mm |
| Weight | About 596g |
| Remote control | WiFi remote control APP / Bluetooth remote control APP / Web remote control |
| ROS support | X |
Applications
- Programming learning and robotics education
- AI vision experiments (object detection, face detection/tracking, gesture interaction, etc.)
- Embodied intelligence demos (vision + voice + motion + grasping)
- Creative interactive projects and technology research
Manuals
- Tutorial link: http://www.yahboom.net/study/DOGZILLA-Lite
Details

A 15-DOF desktop quadruped with an integrated arm and gripper, built for AI interaction and hands-on robotics learning.

Raspberry Pi CM5 computing, camera, microphone, and speaker are integrated for vision, voice, and multimodal interaction.

Multimodal features connect perception, decision-making, and execution—including visual understanding and arm-based handling.

Core capabilities include bionic action groups, multiple vision applications, real-time control, and guided learning resources.

Skip CM5 details here if already covered above; use this graphic as supporting technical context.

Language, voice, and visual AI modes enable text Q&A, voice interaction, and camera-based understanding.

Voice commands and scene understanding work together to trigger actions and responses in real time.

Autonomous-style demos include object selection, color line tracking, and command execution during tasks.

The 3DOF arm and gripper support app-controlled grasping, vision-assisted handling, and obstacle clearing tasks.

A broad set of vision demos covers object/target detection, tracking, gestures, and other camera-based interactions.

Gesture-like inputs such as palm control, skeleton recognition, and face tracking enable playful interactive behaviors.

Extra camera demos include color following, QR-triggered motion, and multi-gesture control options.

A side-by-side comparison helps choose the right DOGZILLA model based on computing, sensors, and functions.

Built-in bionic action groups deliver lifelike moves such as handshake, sit-down, stretching, and more.

The pre-installed GUI brings many features into one place for quick testing, demos, and configuration.

Step-by-step video tutorials support setup, programming, and feature exploration from beginner to advanced use.

The front IPS screen adds expressive feedback with 35 dynamic expressions for more engaging interaction.

Utility functions include recording, posture/servo telemetry, battery monitoring, and multi-robot control performance.

The DOGZILLA-Lite inverse kinematics algorithm supports gait planning with obstacle crossing, walking, crawling, and omnidirectional movement modes.

DOGZILLA-Lite supports cross-platform remote control via Bluetooth and WiFi mobile apps as well as computer web control.

DOGZILLA-Lite supports a teaching mode for manually moving a leg and recording joint angles for movement playback.

DOGZILLA-Lite uses a Raspberry Pi CM5 module with an ESP32 driver board to manage servos and the IMU, with support for peripherals like a camera, SD card, screen, microphone, and speaker.

DOGZILLA-Lite includes a 3-DOF rear robotic arm with an end gripper rated for a 14.4–53.6 mm gripping range for picking up small objects.

DOGZILLA-Lite combines an all-metal quadruped body with a 6-axis IMU, 2500mAh battery, and serial bus servo structure for easier assembly and maintenance.

DOGZILLA-Lite comes with an organized set of folders for AI interaction resources, large model interaction, tutorial videos, and support materials.

DOGZILLA-Lite product parameters list overall dimensions in millimeters along with key component and power details for setup planning.

The DOGZILLA-Lite kit comes with an aluminum box, charging and display cables, a Type‑C USB hub, TF card, tools, and an instruction manual.
