DOGZILLA-Lite 15DOF AI Bionic Dog Robot Dog w/ Raspberry Pi Main Controller 5MP Camera 3DOF Robot Arm

DOGZILLA-Lite 15DOF AI Bionic Dog Robot Dog w/ Raspberry Pi Main Controller 5MP Camera 3DOF Robot Arm

Advantages:
* AI Large-Model Robot Dog
* Embodied Intelligence
* Multi-modal Large Language Model
* Desktop-Grade Bionic Robot Dog
* 15 Degrees of Freedom (including robotic arm)
* Powered by Raspberry Pi Main Controller
* AI Visual Interaction

Description:
DOGZILLA-Lite is a 15-degree-of-freedom desktop-class AI bionic robot dog, equipped with an integrated robotic arm and gripper on its back for flexible object-grasping tasks. It features a built-in Raspberry Pi module that supports various AI vision functions such as facial recognition and object detection. Each joint is driven by 2.3kg·cm serial bus servos, enabling omnidirectional movement, six-axis posture control, posture stabilization, and multiple gait patterns. Powered by a cutting-edge AI large model, DOGZILLA-Lite not only supports intelligent voice interaction and scene understanding, but can also combine arm manipulation with body motion to perform complex embodied intelligence tasks — making it an ideal platform for exploring AI and robotics technology.

Features:
Opening a New Chapter in Embodied Intelligence
* It can perceive, decide, and act.
* This is a thinking, bionic robot dog — your intelligent embodied companion!
* DOGZILLA-Lite can see clearly, understand speech, and even think. Equipped with a 5MP camera, high-sensitivity MEMS microphones, and an enclosed cavity speaker, and powered by the advanced iFLYTEK Spark Large Model, it deeply integrates vision, hearing, and semantic understanding. It can not only accurately recognize objects in its environment, but also autonomously plan task paths and flexibly respond to changes — truly achieving seamless collaboration between hardware and AI.

Multi-modal Large Model Capabilities
* Text-to-Image and Image-to-Text Generation
* Intelligent Voice Interaction
* Intelligent Visual Interaction

Smart Robotic Arm with End-Effector Gripper
* It supports vision-based object positioning and autonomous task step planning.

Built-in Bionic Motion Sets
* Realistically simulates dog-like postures and expressions.

Over 10 AI Vision Applications
* It supports line following, object tracking and following, gesture recognition (contactless control) and intelligent material handling.

Enthusiast-Grade Hardware Configuration
* Full Metal Body
* 15 Degrees of Freedom
* 6-Axis IMU Sensor
* 2500mAh Lithium Battery
* 5MP Camera

Real-Time Video Transmission at Your Fingertips
* It includes Bluetooth/Wi-Fi remote control app for seamless interaction.

Professional After-Sales Technical Support
* Comprehensive service and support to help you resolve any technical issues with ease.

Three Types of AI Large Models
1. Large Language Model (LLM): DOGZILLA-Lite integrates in real time with the iFLYTEK Spark large language model. It not only understands text commands but also responds intelligently and flexibly. It supports text generation, language translation, Q&A interaction, and knowledge summarization.
2. Speech Model: Equipped with high-sensitivity microphones and speakers, DOGZILLA-Lite supports real-time voice-to-text and text-to-voice conversion. With the Spark model connected, it can listen, speak, and engage in intelligent voice interaction.
3. Vision Model: With a built-in 5MP camera, DOGZILLA-Lite can analyze and understand visual input, recognize objects accurately, and respond with text or speech feedback.

Multi-modal Large Model Applications
* Voice Q&A: DOGZILLA-Lite supports free-form voice conversation. With iFLYTEK Spark, it recognizes spoken input and provides meaningful answers.
* Voice Command Execution: It processes spoken commands, interprets intent using the large model, and performs corresponding actions.
* Scene Understanding: Using its vision model, DOGZILLA-Lite can comprehend the scene within its field of view and respond with relevant text or speech.
* Text-to-Image Multi-modal Application: DOGZILLA-Lite can generate high-quality images based on spoken descriptions using multi-modal AI capabilities.

Embodied Intelligence New Experience
* Tracking Kicking: Through the semantic understanding of large language models, DOGZILLA-Lite can lock onto targets based on instructions, adjust its body posture in real time, and accurately complete the tracking kicking action.
* Autonomous Navigation: With the semantic understanding of large language models, DOGZILLA-Lite can accurately recognize and track lines of various colors in real time.
* Embodied Intelligent Robot Dog: DOGZILLA-Lite can hear user commands and analyze the meaning of the instructions through large models. According to different commands, combined with visual recognition content, it responds with different feedback.

Robotic Arm Applications
* APP Remote Control for Handling: The mobile app can not only control DOGZILLA-Lite's free movement but also control the robotic arm's motion to complete tasks of grasping and transporting objects.
* Visual Recognition for Handling: Through AI vision, DOGZILLA-Lite can recognize specified items and grasp and transport them.
* Line-following Automatic Obstacle Clearing: Using AI vision, DOGZILLA-Lite can identify obstacles on its path during autonomous line-following, and use the robotic arm to move and remove the obstacles.
* Embodied Intelligent Handling Control: Through large language models, users can control DOGZILLA-Lite via voice commands to move and place corresponding objects at designated locations.

AI Vision Recognition
* 3D Object Recognition: DOGZILLA-Lite can detect and box three-dimensional objects such as cups, chairs, shoes and cameras.
* Face Detection: DOGZILLA-Lite can detect faces and highlight them with bounding boxes, while also displaying detection confidence.
* License Plate Recognition: DOGZILLA-Lite supports recognizing license plates and displaying the recognition results within the image.
* Object Detection: DOGZILLA-Lite supports detecting common objects.
* Brush Drawing: DOGZILLA-Lite allows drawing on the screen using the fingertip.

AI Vision Fun Interaction
* Air Control: DOGZILLA-Lite calculates fingertip distance by recognizing the palm to perform stand-up and squat actions.
* Human Skeleton Recognition: DOGZILLA-Lite supports recognizing human skeletal posture to imitate standing up and squatting movements.
* Face Tracking: DOGZILLA-Lite can track faces. As the face position changes, the robot dog dynamically adjusts its position accordingly.
* Color Following: DOGZILLA-Lite can track a set target color and execute turning left, turning right, or moving forward based on the color.
* QR Code Motion Control: DOGZILLA-Lite recognizes QR codes and performs corresponding forward, backward, left turn, or right turn actions.
* Face Mask: DOGZILLA-Lite can overlay virtual masks on faces and perform corresponding actions based on the head's position in the image.
* Gesture Following: DOGZILLA-Lite can recognize gesture movements—gesture up makes the robot dog stand up, gesture down makes it squat, and left/right gestures make it move left or right accordingly.
* Gesture Control: DOGZILLA-Lite can recognize gestures such as 1/2/3/4/5/6 fingers, Good, OK and fist, and perform corresponding actions based on the gestures.

Built-in 20 Sets of Bionic Action Groups
* All bionic actions simulate a real puppy in a 1:1 ratio.
* It can shake hands, lie down, mark territory by scent, and stretch.

Pre-installed GUI Program
* The factory pre-installs our self-developed GUI program, which includes over 40 feature-rich functions, integrating most core features. Users can quickly experience functional examples right out of the box, making it easy even for beginners to get started.

Lifelike Dynamic Expressions
* It includes 35 types of dynamic expressions and supports custom dynamic expressions, making every interaction more vivid and interesting.

More Features
* Audio recording and playback
* Video recording and playback
* Posture angle acquisition
* Battery level reading
* Servo angle control
* Height control

Inverse Kinematics Algorithm
* Gait planning, freely adjustable
* Obstacle-Crossing Gait: Always keep three paws touching the ground while lifting one paw
* Walking Gait: Alternately switch contact between two diagonal paws.
* Crawling Forward Gait: Similar to the walking gait, two diagonal paws slowly move forward.
* Omnidirectional Movement Gait: Diagonal pairs of paws alternately move sideways.

Cross-Platform Connected Remote Control
* Bluetooth remote control app (for iOS/Android)
* WiFi remote control app (for iOS/Android)
* PC web control

Support for Teaching Mode
* DOGZILLA-Lite supports teaching functionality, allowing manual control of individual leg movements, recording joint angles, and saving them as files for quick action recording and playback.

Integrated Raspberry Pi CM4 Module
* DOGZILLA-Lite features a built-in AI module based on the Raspberry Pi CM4 platform, equipped with a 2.0-inch IPS color display, 4 programmable buttons, a 5-megapixel camera, digital microphones, and speaker. It offers excellent visual and voice interaction capabilities, enabling image recognition, face detection, voice dialogue, and more.

Raspberry Pi CM4 Specifications
* Model: Raspberry Pi CM4
* CPU: BCM2711 SoC, single-core Cortex-A72 CPU for ARM at 1.5GHz
* GPU: VideoCore VI
* Memory: 2GB
* Codec Support: H.265 (HEVC) 4Kp60 Video Codec; H.264 (AVC) 1080p60 Video Codec and 1080p30 Video Codec

Multi-Master Control Collaborative Operation
* DOGZILLA-Lite adopts a multi-master collaborative architecture. The Raspberry Pi CM4 module acts as the master controller, responsible for computing tasks such as image and voice processing. The ESP32 serves as the secondary controller, managing the robot's power, driving the servos, and executing core motion control algorithms. Through this multi-master collaboration mode, DOGZILLA-Lite achieves more efficient task processing and resource utilization.

Robotic Arm & End-Effector Gripper
* DOGZILLA-Lite is equipped with a 3DOF robotic arm mounted on its back. By combining visual recognition, it can intelligently identify, precisely grasp, and transport various objects. The end-effector gripper features current protection feedback to safeguard both the gripped objects and the gripper's servo motor.
* The gripping range of the gripper is 14.4 to 53.6mm.

High-Quality Materials
* Full aluminum alloy body
* ABS wear-resistant foot
* Shock-absorbing silicone foot rods
* 2500mAh large capacity battery
* Joint servo angle feedback
* 6-axis IMU posture sensor

Over 100 Systematic Detailed Courses
* From beginner to advanced — a complete learning path for both hobbyists and engineers.
* Helping users seamlessly learn machine intelligence.
* Tutorial Link: http://www.yahboom.net/study/DOGZILLA-Lite

All Course Codes Fully Open Source
* The materials can be browsed online via web pages and also support packaged downloads. Our company has purchased Baidu Netdisk high-speed sharing service, allowing ultra-fast downloads even without a Baidu membership.
* Professional after-sales technical support is provided to ensure a smooth learning journey.

Product Specifications:
* Product Name: DOGZILLA-Lite AI Large-Model Robot Dog
* Dimensions (Powered On): 240.47 × 142.93 × 168.45mm
* Dimensions (Powered Off): 250.58 × 142.99 × 111.61mm
* Battery: 7.4V 2500mAh Lithium Battery
* Material (Body): 1mm Aviation Aluminum Alloy
* Material (Feet): ABS Plastic
* Architecture: Upper Computer: Raspberry Pi CM4 Module; Lower Controller: ESP32
* Servo Model: 15 × Serial Bus Servos
* Servo Torque: 2.3kg·cm
* Gripper Range: 14.4-53.6mm
* Screen: 320 × 240 Full-Color IPS
* Camera: 5MP OV5647
* Speaker: 2W 8Ω Enclosed Cavity Speaker
* Microphone: Dual MEMS Digital Microphones
* Storage: 32GB TF Card
* Charging Port: Type-C 5V
* Charging Time: 2.5 hours
* Battery Life: 3 hours
* Weight: Approx. 575g  

Packing List:
* Assembled Robot Dog × 1
* Aluminum Carrying Case × 1
* Type-C Charging Cable × 1
* Micro HDMI-Compatible to HDMI-Compatible Cable × 1
* Type-C USB Hub × 1
* 32GB TF Card × 1
* Wooden Blocks × 3
* EVA Small Balls (one each in red, blue, green) × 3
* Screw Pack + Screwdriver × 1 Set
* User Manual × 1
    
Packaging Details:
* Weight: 1.73kg

Note:
* Servos are delicate components. Do not apply excessive force or press on them.