As2-W EDU Smart — wheeled-leg quadruped with 100 TOPS AI compute, depth camera, and charging pad for autonomous docking

What's in This Configuration

The Unitree As2-W EDU Smart is the upgraded development platform in the As2-W EDU series — a compact yet extraordinarily capable wheeled-leg quadruped robot designed for advanced education, research, and AI-driven prototyping. It combines the high-speed efficiency of wheels with the obstacle-overcoming capability of articulated legs, automatically adapting to stairs, gravel, and steep slopes. Built around 16 industrial-grade joint motors with 95 N·m peak torque, it climbs obstacles up to 80 cm high and slopes up to 45°. The 7-inch oversized wheels (178 mm diameter, 50 mm width) deliver smooth rolling on flat surfaces. The robot weighs approximately 25 kg with battery and carries a continuous walking payload of 16 kg — with a maximum standing load of 150 kg. Powered by a 15,000 mAh (648 Wh) lithium-ion battery, it achieves over 3 hours of continuous operation unloaded with a range exceeding 30 km. The included charging pad enables autonomous docking and contactless recharging — essential for long-duration autonomous missions. The upgraded 100 TOPS AI expansion dock delivers 2.5× the inference performance of the Standard configuration, enabling real-time processing of large AI models, multi-task perception, and complex decision-making pipelines. A depth camera is now included for 3D environment mapping, obstacle avoidance, and spatial reasoning. The industrial-grade 64–128-line ultra-wide-angle LiDAR and HD front camera provide rich environmental perception. The ISS 3.0 Intelligent Side-follow System delivers centimeter-level positioning for hands-free companion operation. IP54 protection and −20°C to 50°C operating range ensure reliability in demanding conditions. Full Wi-Fi 6, Bluetooth 5.2, speaker, array microphone, and front lighting enable rich interaction. A handheld remote controller is included for manual operation. The UniStore Application Platform and Smart OTA Updates keep capabilities growing.

Wherever It Goes, Conquers Every Terrain

The As2-W EDU Smart combines the high-speed efficiency of wheels with the obstacle-overcoming capability of legs, automatically adapting to stairs, gravel, and steep slopes. It traverses complex terrain with ease, climbing obstacles up to 80 cm high and slopes up to 45°. The 7-inch oversized wheels (178 mm diameter, 50 mm width) provide smooth, fast rolling on flat and paved surfaces, while the articulated leg suspension absorbs shocks and maintains traction on uneven ground. The extensive joint range of motion — body: −45° to 45°, thigh: −115° to 200°, shank: −159° to −47° — enables natural, animal-like gait adaptation for maximum stability and maneuverability across any surface.

Open Platform, Empowering Innovation: developer-friendly SDK and APIs support large AI models for embodied AI interaction and autonomous decision-making.

100 TOPS AI Expansion Dock — Power for Large-Model Inference

The As2-W EDU Smart ships with an upgraded High-Performance Expansion Dock delivering 100 TOPS of AI inference performance — a 2.5× leap over the EDU Standard's 40 TOPS. This transforms the robot into a high-performance autonomous intelligent agent capable of real-time large-model perception, multi-task decision-making, and complex scene understanding. The 100 TOPS compute platform runs the full NVIDIA software stack including CUDA, TensorRT, and JetPack, enabling rapid deployment of large deep learning models for object detection, semantic segmentation, 3D reconstruction, visual SLAM, and reinforcement learning. Unitree provides technical support for AI algorithm development, including sample code, model repositories, and integration guides. Combined with the on-board LiDAR, HD camera, and depth camera, the 100 TOPS platform unlocks research in autonomous navigation, terrain understanding, dynamic obstacle avoidance, multi-modal sensor fusion, and embodied intelligence — all running on-board without cloud dependency.

1× Depth Camera — See the World in 3D

Unlike the EDU Standard where depth perception is optional, the As2-W EDU Smart includes a depth camera as standard equipment. This adds a critical third dimension to the robot's perception stack, enabling 3D environment mapping, precise obstacle avoidance, spatial reasoning, and depth-based visual SLAM. The depth camera works in concert with the ultra-wide-angle LiDAR and HD front camera to build rich, dense 3D representations of the surrounding environment. Researchers can explore point-cloud processing, volumetric scene understanding, depth-guided navigation, and 3D object recognition — all powered by the 100 TOPS AI compute platform. The depth camera is particularly valuable for indoor navigation, cluttered environment traversal, and close-range manipulation planning where LiDAR alone may have blind spots.

Powerful Performance, Long Endurance

The As2-W EDU Smart is driven by 16 industrial-grade joint modules delivering ~95 N·m peak torque each. Low-inertia, high-speed inner-rotor permanent magnet synchronous motors (PMSM) provide exceptional response speed and heat dissipation. Industrial-grade crossed-roller bearings on every joint ensure high precision and high load capacity under continuous operation. The 15,000 mAh (648 Wh) lithium-ion battery powers over 3 hours of unloaded cruising with a range exceeding 30 km. Even under a 16 kg continuous walking load, the robot maintains over 2 hours of operation covering approximately 25 km. The localized air-cooling system keeps joint temperatures in check during high-intensity missions. With a maximum standing load of 150 kg, the As2-W EDU Smart can support heavy equipment, sensor payloads, or experimental apparatus in stationary applications.

Autonomous Side-Follow, Ready to Go With You

The ISS 3.0 Intelligent Side-follow System delivers centimeter-level positioning accuracy and stable tracking, allowing the As2-W EDU Smart to follow its operator autonomously through complex environments. Whether carrying backpacks and supplies for outdoor exploration, transporting tools across a campus, or accompanying personnel during inspection rounds, the robot maintains a safe, consistent distance and adapts its speed to the operator's pace. The system leverages the on-board HD camera, depth camera, LiDAR, and 100 TOPS AI compute for real-time person detection, 3D path planning, and obstacle avoidance — no external beacons or GPS required.

Autonomous Charging Pad — Always Ready

The As2-W EDU Smart includes a charging pad for autonomous docking and contactless recharging. The robot can navigate back to its charging station at the end of a mission, align itself precisely using LiDAR, depth camera, and visual fiducial markers, and begin recharging without human intervention. This enables fully autonomous long-duration deployments — the robot can patrol, inspect, or collect data continuously, returning to base only when battery levels require topping up. Combined with the fast charger (50.4 V, 7.4 A), downtime is minimized, and operational availability is maximized.

Reliable and Robust, Even in Harsh Conditions

Built for real-world deployment, the As2-W EDU Smart carries an IP54 dust and water resistance rating, withstanding rain, dust, and wet terrain without compromise. It operates reliably across a wide temperature range from −20°C to 50°C, making it suitable for everything from cold-weather outdoor research to hot industrial floors. The aluminum alloy + high-strength engineering plastic chassis balances lightweight agility with structural durability. Dual encoders on every joint provide closed-loop precision and fault detection. The robot is engineered for 24/7 mission readiness — whether patrolling perimeters, collecting field data, or supporting advanced academic research.

Triple-Sensor Perception — LiDAR, Depth & Vision

The As2-W EDU Smart is equipped with a comprehensive triple-sensor suite for full 3D situational awareness. The industrial-grade 64–128-line ultra-wide-angle LiDAR captures dense 3D point clouds of the surrounding environment for SLAM mapping, obstacle detection, and path planning. The HD front camera delivers high-definition video streaming and visual recognition for object detection, person tracking, and terrain analysis. The included depth camera adds precise 3D depth measurements for close-range obstacle avoidance, stair detection, and spatial mapping. Together with the ISS 3.0 side-follow module and 100 TOPS AI processing, these sensors enable autonomous navigation, intelligent following, 3D scene reconstruction, and safe operation in dynamic, unstructured environments.

Comprehensive sensor and compute architecture: LiDAR, HD camera, depth camera, 100 TOPS AI compute, 8-core CPU, high-power motors, smart battery, and rich external interfaces

Compact Foldable Design — Go Anywhere

Despite its extraordinary capability, the As2-W EDU Smart is remarkably compact. In standing configuration, it measures 721 × 493 × 521 mm — small enough to navigate doorways, corridors, and tight spaces. When folded, it compacts to just 768 × 602 × 211 mm, fitting easily into vehicle trunks, equipment cases, or storage lockers for transport and deployment. At approximately 25 kg with battery, it is light enough for two people to lift and position, yet heavy-duty enough to carry substantial payloads and withstand demanding operational conditions.

Open Platform, Empowering Innovation

Unitree provides a mature, developer-friendly SDK and open APIs for the As2-W EDU Smart platform. The 8-core high-performance CPU runs the real-time control stack, while the 100 TOPS AI expansion dock handles large-model perception and decision-making. External interfaces — Gigabit Ethernet ×1, SBUS ×1, and BAT power output — enable integration with custom payloads, sensors, and compute modules. The UniStore Application Platform offers a growing ecosystem of apps and capabilities. Smart OTA Updates ensure the robot continuously improves with new features and optimizations. The platform supports custom development with provided robot models and simulation interfaces, including compatibility with Isaac Sim and other simulation environments for rapid algorithm prototyping and reinforcement learning research. Unitree provides excellent technical support services, comprehensive development manuals, and ecosystem support to accelerate your projects from concept to deployment.

Upgrade Path — Grow With Your Research

The As2-W EDU Smart is the mid-tier platform in the EDU series. Upgrade your capabilities as your research and projects evolve:

• AI Compute — Already at 100 TOPS. This matches the compute in Plus and Ultimate configurations — no upgrade needed for AI performance.
• Depth Perception — Already at 1× depth camera. Upgrade to 3× depth cameras (Ultimate) for 360° spatial awareness and multi-view fusion research.
• Navigation LiDAR — Upgrade to Livox Mid-360 LiDAR (Plus, Ultimate) with technical support for navigation algorithms and advanced SLAM.
• Multi-Camera Perception — Upgrade to 3× depth cameras (Ultimate) for full spherical 3D perception and multi-view depth fusion.
• Charging Infrastructure — Charging pad already included. Add additional pads for multi-zone autonomous operation.

Designed For

Technical Specifications

What's in the Box

• Unitree As2-W EDU Smart Wheeled-Leg Quadruped Robot
• 100 TOPS High-Performance AI Expansion Dock (pre-installed)
• 1× Depth Camera Module (pre-installed)
• 15,000 mAh long-life lithium-ion battery (pre-installed)
• Fast charger (50.4 V, 7.4 A)
• Charging pad for autonomous docking
• Handheld remote controller
• Ultra-wide-angle LiDAR module (pre-installed)
• HD front camera module (pre-installed)
• Front lighting system (pre-installed)
• Speaker & array microphone module (pre-installed)
• ISS 3.0 Intelligent Side-follow System (software-enabled)
• Quick-start guide & safety manual
• Secondary development documentation access

As2-W EDU Configuration Guide

The As2-W EDU series shares the same 16-DOF wheeled-leg platform — the difference is the AI compute, sensors, and development tools included. Choose the one that matches your research goals:

* Specifications are based on manufacturer data and may vary slightly by production batch. Optional accessories sold separately.
* GPS and 4G functions are supported only by the Mainland China version and are disabled by default; enabled after user authorization.
* The above data was obtained under laboratory conditions. Actual performance may vary based on usage conditions.
Unitree is a registered trademark of Unitree Robotics. Images are for reference; actual product appearance may differ.
Users must maintain a sufficient safety distance. This is a civilian robot — comply with local laws and regulations.