Robots can see the world.
But true dexterity requires touch.
At @TexavieInc , our 5k1n tactile intelligence platform captures not only precise hand pose, but also rich multimodal tactile interaction across the entire hand.
In this simple fun demo, our 5k1n smart tactile gloves are used to simply manipulate a tennis ball while capturing:
• Full-hand unobstructed pose intelligence
• Dense tactile sensing on fingers and palm
• Contact-rich soft object manipulation
• Egocentric multimodal data
• Real-world interaction dynamics
From grip adaptation to rolling contact points and dynamic force distribution, these interactions generate valuable training data and are basis for Tactile Foundation Model for the next generation of humanoid robots and physical AI systems.
Vision alone is insufficient for dexterous robotics.
Touch is the missing layer.
5k1n is designed to help robots learn how humans naturally interact with the physical world.
More to come! Get in touch to learn more.
#Robotics #PhysicalAI #Humanoids #EmbodiedAI #TactileSensing #DexterousManipulation #AI #MachineLearning #RobotLearning #SmartTextiles #Texavie #5k1n #TactileFoundation
A tennis ball becomes rich tactile intelligence data with 5k1n.
Capturing pose, touch, force, and contact-rich manipulation for the next generation of physical AI.
A tennis ball becomes rich tactile intelligence data with 5k1n.
Capturing pose, touch, force, and contact-rich manipulation for the next generation of physical AI.
Robots can see the world.
But true dexterity requires touch.
At @TexavieInc , our 5k1n tactile intelligence platform captures not only precise hand pose, but also rich multimodal tactile interaction across the entire hand.
In this simple fun demo, our 5k1n smart tactile gloves are used to simply manipulate a tennis ball while capturing:
• Full-hand unobstructed pose intelligence
• Dense tactile sensing on fingers and palm
• Contact-rich soft object manipulation
• Egocentric multimodal data
• Real-world interaction dynamics
From grip adaptation to rolling contact points and dynamic force distribution, these interactions generate valuable training data and are basis for Tactile Foundation Model for the next generation of humanoid robots and physical AI systems.
Vision alone is insufficient for dexterous robotics.
Touch is the missing layer.
5k1n is designed to help robots learn how humans naturally interact with the physical world.
More to come! Get in touch to learn more.
#Robotics #PhysicalAI #Humanoids #EmbodiedAI #TactileSensing #DexterousManipulation #AI #MachineLearning #RobotLearning #SmartTextiles #Texavie #5k1n #TactileFoundation
Last week, American/Israeli fighter jets targeted Iranian Pistachio Company warehouses. The Resnick family's Beverly Hills-based pistachio empire stands to gain. You won't believe the backstory ⤵️
@pliang279 Great work! You should use @TexavieInc 5k1n tactile gloves to get exceptional egocentric tactile data. Much higher in quality and resolution. Let’s chat. 🙏
@altium accounting is one of the worst. You ask them to set up a specific payment and when they forget they ask you to pay for their mistake. One of the worst!
🚀 Introducing TACTILEFOUNDATION-0.1
Robots can see.
They still don’t understand touch and contact.
Vision-only systems break down when objects slip, deform, require force control, or involve tools.
That’s where real-world manipulation actually begins: touch and manipulation.
Today we’re excited to release TactileFoundation-0.1 — a real-world tactile dataset and learning model for physical AI, powered by our 5k1n tactile technology.
TactileFoundation-0.1 includes:
• Full-hand tactile pressures
• Force and slip signals
• High-precision hand pose
• Egocentric vision
• Time-synchronized multimodal data
• Human dexterity demonstrations in real environments
This dataset is designed to accelerate:
→ Contact-aware dexterity and manipulation
→ Tool use and soft object handling
→ Sim-to-real robustness
→ Foundation models for robotics
Language models scaled on text.
Vision models scaled on images.
Robots will scale on TactileFoundation-0.1.
We’re onboarding early partners across humanoid robotics, embodied AI, and manipulation learning to get access to TactileFoundation-0.1.
If this aligns with what you’re building, feel free to reach out or comment below.