Solving the ‘Hand Problem’: Proception Secures Funding for High-Dexterity Grippers
Proception has settled a trade secret suit with Tesla and secured $11M in funding to solve the robotic hand problem. Their approach uses innovative data collection to tackle the most difficult challenge in dexterity.
Robotic hands remain one of the most significant engineering hurdles in the field of automation. While humanoid torsos and legs have achieved record-breaking speeds—as seen with recent humanoid marathon milestones—the fine motor skills required for human-like manipulation are still lagging. Startup Proception is aiming to change that following an $11 million funding round and the resolution of a high-profile legal dispute with Tesla.
Proception’s focus is on the "hand problem"—specifically, how to collect high-quality training data for tactile manipulation. Many current robots can pick up a box, but few can handle delicate tools, tie knots, or manage the varied textures of a domestic kitchen. Proception uses a unique data collection method that captures the nuance of human touch and translates it into training sets for neural networks.
The settlement with Tesla allows the company to move forward without the shadow of litigation, focusing on commercializing grippers that could revolutionize industries from electronics assembly to elder care. As dexterity improves, we are seeing a shift from robots that simply "move" things to robots that can "work" in the same way humans do, potentially unlocking billions in economic value in the service sector.
Source: TechCrunch
