Hi Product Hunt, we’re Proception.
We’re building ProHand, a 22-DOF tendon-driven robotic hand, and ProGlove, a wearable sensor glove for collecting human manipulation data.
The problem we’re focused on is dexterous manipulation. Robots have become much better at seeing and navigating, but most still interact with the world using simple grippers. That works for structured tasks, but it breaks down when a task needs multi-finger contact, grip adjustment, tactile feedback, or recovery from slip.
We think the bottleneck has two parts:
1. Robot hands need to be much closer to human hands in kinematics and sensing.
2. Robot learning needs better manipulation data, especially real human contact data.
ProHand is our execution platform. It has 22 degrees of freedom, tendon-driven actuation, and skin-like contact sensing.
ProGlove is our data platform. It lets people collect human hand motion and contact data while manipulating real objects, without needing a robot in the loop.
Together, they create a human-to-robot manipulation pipeline: collect human interaction data, then transfer those movements and skills to a robotic hand with similar structure and sensing.
We’re shipping the first batch of ProHand units to robotics researchers and companies.
We’d love feedback from people working on robotics, AI, hardware, teleoperation, tactile sensing, simulation, or human data collection. What would you want to see in a robotic hand and glove platform?
Proception’s robotic hand lets us demonstrate real-time control from forearm EMG. It gives us a physical platform to test how decoded human finger movements translate into robot hand motion.
