Tangent Robotics Raises $4.5 Million to Give Robot Hands Fine Motor Skills

Tangent Robotics has raised $4.5 million in pre-seed funding to develop dexterous robot hands for manufacturing. Fly Ventures and Toyota Ventures co-led the round, with participation from Logos Fund and Sparked Ventures, the New York company announced on September 30.

Key Takeaways

Show all

• Tangent raised $4.5 million to develop and deploy fine manipulation capabilities in manufacturing.
• The Columbia University spinout combines hand hardware, touch sensing and learned motor skills.
• Its three-finger TR02 illustrates a focus on dexterity without reproducing human hand anatomy.

Spun out of Columbia Engineering’s Robotic Manipulation and Mobility Lab, Tangent was founded by Pedro Piacenza, Matei Ciocarlie and Ioannis Kymissis. Its target applications include fitting connectors, seating gaskets, threading and assembling gears.

Tangent describes its focus as a fine motor skill layer between vision-language-action models and low-level controllers: the learned movements needed to execute delicate work. It combines optical touch sensing with robot hands and training methods, aiming eventually to shorten the learning process to days and then hours.

Dexterity without five fingers

The company’s TR02 hand, introduced in June, has three fingers and 10 degrees of freedom. Each fingertip incorporates a Popcorn touch sensor, while an offset palm camera adds visual feedback.

Tangent says the third finger and palm expand on its two-finger TR01 design, allowing more stable grasps, movement of objects within the hand, and simultaneous holding and manipulation. Its examples include holding a jar while opening the lid and moving an object from a fingertip grasp into the palm. The design also tracks an operator’s fingertip movements for collecting training demonstrations.

For humanoid robotics, the relevance is the manipulation problem: reaching an object is only one part of completing an industrial task. Aligning, gripping and adjusting it through contact can determine whether a robot performs useful work. Tangent’s funding will support an attempt to turn those fine movements into deployable factory capabilities.