Smartgloves II: Best-in-class finger capture gloves

Rokoko Smartgloves are professional wireless hand and finger motion capture gloves. They use a proprietary combination of IMU (inertial measurement unit) and EMF (electro-magnetic field) sensor fusion technology - 7 sensors per glove - to capture precise, real-time movement data for every finger and the wrist, with no cameras or markers required.

Designed to be lightweight, wearable, and deployable in any environment, Smartgloves output high-fidelity hand pose data that can be streamed live or recorded and exported. Whether you're animating a character, training a robot, or operating machinery from a distance, Smartgloves give you accurate hand and finger data that you can act on immediately.

Yes. Smartgloves II are designed to integrate with both Rokoko's own ecosystem and third-party pipelines. For users who already use mocap tools from other optical mocap brands (like Optitrack, Qualisys, MOVIN), Rokoko Smartgloves can be added to their pipeline thanks to real-time native plugins for all major 3D tools and software.

Native plugins are available for Optitrack, Qualisys, MOVIN, Unreal Engine, Unity, Blender, Maya, MotionBuilder, and Houdini. Data can also be streamed via standard protocols or exported to CSV, FBX and BVH formats compatible with industry-standard skeleton presets including Human IK and Mixamo.

For robotics and teleoperation workflows, Smartgloves II outputs structured hand pose data that can be consumed by custom pipelines, ROS-based systems, and machine learning frameworks - making them a flexible input device beyond the traditional animation toolchain.

Smartgloves have a wide range of professional applications across creative and technical industries:

Entertainment - Capture nuanced hand and finger performances for animated characters, VFX, and game development. Pair with the Smartsuit Pro II for full-body capture and Face Capture devices for complete performance data.

Robotics - Use hand pose data to teach robotic systems natural, human-like hand and finger movements. Smartgloves provide clean, structured data needed to train robotic hands and dexterous end-effectors. Pair with the Coil Pro for absolute position in space - no drift, no occlusion.

Teleoperation - Operate robotic hands and tools remotely with precision. Smartgloves stream low-latency hand pose data that can be mapped directly onto robotic systems, enabling intuitive, human-in-the-loop control.

Machine learning & data collection - Generate high-quality, hand motion datasets at scale for training AI and ML models, including gesture recognition, dexterous manipulation, and human-robot interaction systems.

Yes. Smartgloves II are designed to integrate with both Rokoko's own ecosystem and third-party pipelines. For users who already use mocap tools from other optical mocap brands (like Optitrack, Qualisys, MOVIN), Rokoko Smartgloves can be added to their pipeline thanks to real-time native plugins for all major 3D tools and software.
Native plugins are available for Optitrack, Qualisys, MOVIN, Unreal Engine, Unity, Blender, Maya, MotionBuilder, and Houdini. Data can also be streamed via standard protocols or exported to CSV, FBX and BVH formats compatible with industry-standard skeleton presets including Human IK and Mixamo.
For traditional animation and content creation workflows, Rokoko Studio is the recommended software. It handles device setup, real-time streaming, recording, filtering, retargeting, and export.

For technical users in robotics, teleoperation, and machine learning, Smartgloves II can also be used without a full Rokoko Studio workflow. Raw sensor data and hand pose output can be accessed directly for integration into custom systems and pipelines.

Smartgloves II delivers meaningful hardware and firmware upgrades across tracking performance, data quality, and durability.

Tracking accuracy has improved significantly. IMU sample rate is 50% faster (250 Hz vs 167 Hz), capturing more detail in fast finger movements. A refined sensor fusion model and the addition of a magnetometer eliminate heading drift and deliver more consistent hand poses. Wrist tracking accuracy has improved 3×, from 1.5mm to 0.47mm at fingertip distance.

Data quality is cleaner out of the box. A new ADC reduces noise pickup by 16×, and TCP-based WiFi streaming replaces UDP for more stable transmission with less packet loss. For ML and robotics workflows, this means less filtering and post-processing — the data is production-ready faster.

Durability and comfort have been overhauled. Reinforced construction and cable strain relief make the gloves far more robust under frequent or high-impact use. 33% lower power consumption delivers 50% longer battery life and eliminates heat buildup during extended sessions.

Setup is faster and more reliable. LED rings provide visual actor identification, an on-device button enables direct control (start or stop recordings), and the TCP WiFi connection is more firewall-friendly and stable than the original.