Image:
PaXini
A humanoid robot featuring advanced multidimensional tactile sensors in its hands for precise manipulation.
Vision-Touch-Language-Action (VTLA).
Height
186
cm
Weight
N/A
kg
Speed
1.0
m/s
Payload
12
kg
Actuators
Modular electric actuators integrated with ITPU-based multi-dimensional tactile sensors for high-precision manipulation.
DoF (Domains of freedom)
21
°
ITPU-based sensors perceive force, pressure, texture, and material properties.
Rapidly learns human production tasks through visual-tactile dual-modal data.
Vision-Touch-Language-Action AI for high generalization across manufacturing tasks.
Telescoping height adaptation to suit diverse industrial and service workspaces.
A humanoid robot featuring advanced multidimensional tactile sensors in its hands for precise manipulation.
Standing 186 cm, Tora-one features 21 degrees of freedom. It utilizes an NVIDIA AGX Orin for a "cerebrum + cerebellum" control architecture and features a fully foldable upper body.
Image:
PaXini
Specializes in hazardous environments and precision manufacturing. Key features include 3D LiDAR, fisheye cameras, and the ability to perform imitation learning based on visual-tactile dual-modal data.
Tora One by PaXini
Actuators
Modular electric actuators integrated with ITPU-based multi-dimensional tactile sensors for high-precision manipulation.
DoF (Domains of freedom)
21
°
Height
186
cm
Speed
1.0
m/s
Weight
N/A
kg
Payload
12
kg
Runtime
N/A
h
OS / AI System
Vision-Touch-Language-Action (VTLA).
Uses proprietary ITPU-based multi-dimensional tactile sensors on its hands. Its brain is powered by the VTLA (Vision-Touch-Language-Action) model for strong generalization in complex tasks.
Image:
PaXini
Customers include electronics assembly plants and food-service providers (e.g., automated kiosks). It is also utilized in hazardous zones where human-like dexterity is required to operate existing tools and levers.
Improvements over standard robotic arms include the integration of ITPU-based multi-dimensional tactile sensors and a VTLA (Vision-Touch-Language-Action) model, allowing it to perform tasks like ice cream making with human-like touch.
Learns production tasks through a dual-modal model that fuses vision and touch, allowing it to adopt the "muscle memory" and delicate handling techniques of a human craftsman.
Specifically engineered to replace humans in hazardous zones, its AI prioritizes accident prevention and protective behavior during close-range collaboration.
Equipped with ITPU-based sensors that allow it to "feel" the shape and texture of objects, enabling a human-like touch sensitivity for high-precision manufacturing tasks.
by
Unitree Robotics
A general-purpose, full-size humanoid robot focused on affordability and high-performance locomotion.
by
Galbot
A full-sized, electric-powered bipedal humanoid robot focused on autonomous task execution.
by
AgiBot
A general-purpose humanoid robot platform designed for scalable industrial applications and AI development.
by
SwitchBot
A multitask household humanoid robot utilizing VLA models for complex chores like laundry and dishwashing in real home environments.
