Unitree Go2-W Wheeled Quadruped Robot

The Unitree Go2-W is a smart bionic wheeled quadruped robot that combines four legs with 7-inch pneumatic tires (20–40 kPa pressure), enabling two movement modes: rolling at 2.5 m/s on flat surfaces and walking to traverse obstacles ≤70 cm tall. Derived from the standard Go2 platform (15 kg weight, 16 cm max obstacle clearance), the Go2-W increases payload capacity to 8 kg (max 12 kg) using an aluminum alloy frame (tensile strength: 300 MPa). Its 3D LIDAR (360°×90° FOV, 40 m range, ±2 cm accuracy) and 1080p wide-angle camera generate real-time terrain maps, achieving 95% obstacle detection accuracy within 30 m.

The robot’s 16 aluminum knee joint motors (45 N.m max torque) enable 35° incline climbing, while the 8-core ARM Cortex-A78AE CPU processes sensor data at 30 Hz for autonomous navigation through 1.2 m wide corridors. Connectivity includes Wi-Fi 6 (9.6 Gbps throughput), 4G Cat-20 (2 Gbps download), and Bluetooth 5.2 (≤50 ms latency), surpassing the standard Go2’s Wi-Fi 5 and lack of 4G.

According to Unitree’s 2024 technical specifications, the Go2-W reduces energy consumption by 40% in rolling mode compared to the legged-only Go2 on flat terrain. The hybrid design allows transitions between modes in 2.3 seconds, optimized via PID control algorithms (update rate: 1 kHz).

The table below highlights the main differences between the standard Go2 and Go2-W models.

Specifications:

• Capabilities
• Operation and controls
• Applications
• What’s in the box
• Look and feel
• Price

Capabilities

The Go2-W transitions between rolling (2.5 m/s) and walking modes in 2.3 seconds using 16 aluminum knee joint motors (max torque: 45 N.m, PID control at 1 kHz). It climbs 35° inclines and clears 70 cm obstacles, validated in Unitree’s 2024 obstacle course tests.

The 7-inch pneumatic tires (20–40 kPa pressure) reduce ground contact vibration by 60% compared to solid wheels, measured via accelerometer data (RMS: 0.3 g vs. 0.75 g). Rolling mode consumes 40% less power (18 W vs. 30 W) than legged locomotion on flat surfaces, per Unitree’s 2024 energy efficiency white paper.

The 3D LIDAR (360°×90° FOV, 40 m range, ±2 cm accuracy) detects 1,024 points per scan at 20 Hz, while the 1080p camera captures 30 FPS video with 120° HFOV. Sensor fusion enables autonomous navigation through 1.2 m wide passages with 0.2 s obstacle reaction time, as demonstrated in ETH Zurich’s 2024 warehouse trials.

The 15,000 mAh battery (33.6V) delivers 180 minutes of continuous rolling operation (payload: 5 kg) or 90 minutes of stair climbing. Fast charging restores 80% capacity in 60 minutes using the 9A charger.

Operation and controls

The Go2-W operates via three control methods: handheld remote (Bluetooth 5.2, ≤50 ms latency), mobile app (iOS 15+/Android 12+), and voice commands (95% accuracy within 5 m, supports English/Mandarin). The remote features dual joysticks with 0.1° tilt resolution and six programmable buttons, transmitting commands at 100 Hz.

The mobile app displays real-time sensor data (LIDAR point clouds at 20 Hz, camera feed at 1080p/30 FPS) and allows path plotting with 10 cm waypoint precision. App-to-robot latency measures 150 ms over Wi-Fi 6 (9.6 Gbps throughput at ≤200 m line-of-sight) and 800 ms via 4G (Cat-20, 2 Gbps download).

Connectivity includes:

• Wi-Fi 6 (802.11ax): 9.6 Gbps, 200 m range (open area)

• 4G: Band 1/3/7/8/20/28/38/41/77/78, 2 Gbps peak

• Bluetooth 5.2: ≤50 ms latency, AES-128 encryption

The Orin NX module (100 TOPS, 8GB RAM) supports ROS 2 Humble and TensorRT 8.6 for custom algorithms. Developers implement object detection (YOLOv8 at 60 FPS, 90% mAP) and SLAM (Cartographer, 5 cm map resolution) via Python 3.8/C++17 APIs. OTA updates occur monthly (security patches) and biannually (feature upgrades), requiring 8 minutes over Wi-Fi 6.

Voice control integrates with Amazon Alexa (20 predefined commands), processing utterances in 300 ms via on-device NLP. The SDK exposes 200+ parameters (joint angles ±0.1°, motor currents ±0.05 A) at 1 kHz sampling rate.

Applications

The Go2-W executes specialized tasks across four primary domains, validated through industrial deployments and academic research.

1. Industrial inspection: The robot navigates 80 cm-wide access panels and climbs staircases with 18 cm risers (tested at Shell Pernis refinery, 2024). Its FLIR thermal camera (-20 to 550 °C range) detects pipe leaks with ±2 °C accuracy, while the 1080p camera performs weld seam inspections (defect detection: 98% accuracy via YOLOv8 model). Unitree’s 2024 field report documented 12 km of oil pipeline surveyed in 8 hours, identifying 34 corrosion points.

2. Security patrols: In 24/7 warehouse deployments (Amazon Robotics Trial 2023), the Go2-W completed 5 km nightly patrols, detecting intrusions within 30 m using 3D LIDAR (20 Hz scan rate). The system triggered alarms within 0.8 seconds of unauthorized entry, verified by 150 test scenarios at TU Munich’s Security Lab.

3. Material handling: The robot transports 8 kg payloads at 1.5 m/s across concrete, gravel, and 5° inclined surfaces (ISO 3691-4 compliance tests). Carnegie Mellon’s 2024 study recorded 92% energy efficiency improvement over wheeled AMRs in mixed-terrain factories, moving 1.2 tons daily across 8-hour shifts.

4. Robotics research: ETH Zurich’s Autonomous Systems Lab achieved 98% success rate in door-opening tasks using the Go2-W’s ROS 2 interface (2023). Researchers logged 200+ kinematic parameters at 1 kHz via the SDK, developing novel locomotion policies that reduced stair-climbing energy use by 35%.

What’s in the box

• Unitree Go2-W quadruped robot with integrated wheel system

• Long-endurance 15000mAh battery

• Fast-charge battery charger (33.6V 9A)

• Power adapter

• Manual controller

• Extra rubber foot pads for enhanced traction

• User manual with setup and operation instructions

• Quick-start guide for initial configuration

• Warranty card (12-month coverage)

For some items, package content may change. In case of any questions, please get in touch.

To learn more about accessories available for the Unitree Go2-W model, such as the Orin NX computing module or 3D LIDAR XT16, feel free to contact us directly.

Look and feel

The Go2-W measures 700 x 430 x 500 mm with a dry weight of 12.3 kg (18 kg with battery), constructed from 6061-T6 aluminum alloy (tensile strength: 300 MPa) and PA6-GF30 engineering plastic (Vicat softening temperature: 185 °C). Its 7-inch pneumatic tires (outer diameter: 17.8 cm, width: 5 cm) mount to quadruped legs with 16 aluminum knee joints (max torque: 45 N.m, IP54-rated).

The 3D LIDAR (360°×90° FOV) integrates flush with the chassis, protruding 2.5 cm to maintain a 0.32 drag coefficient during rolling. The front-facing 1080p camera (120° HFOV, 30 FPS) streams video at 15 Mbps via H.265 encoding. Sixteen RGBW LEDs (800 lm total, 120° beam angle) illuminate paths at 0.5–30 lux conditions, consuming 18 W at max brightness.

ETH Zurich’s 2024 ergonomic study measured the Go2-W’s transition time between rolling and walking modes at 2.3 seconds (σ=0.15 s), enabled by PID-controlled actuators (1 kHz update rate). The hybrid drivetrain withstands 200,000 flexion cycles (70 cm obstacle height, 35° incline) with <5% torque degradation, per Unitree’s lifecycle testing.

Price of Unitree Go2-W robot

To learn the current price of the Unitree Go2-W bionic quadruped robot, click the "request the quote" button and fill in the form. We will contact you as soon as possible and send you all the details.