The humble human fingernail, often overlooked as a biological feature, turns out to be a surprisingly powerful tool for delicate manipulation — and now, engineers have put that lesson to work in a robot.
Researchers have developed a robotic hand with artificial fingernails embedded in soft, skin-like fingertips, enabling it to pick up playing cards, peel oranges, and separate individual sheets of paper from a stack, researchers describe February 5 in arXiv. The design, they say, outperforms conventional soft-fingertip robots on tasks requiring precise, delicate contact — tasks that have long confounded robotic systems.
The inspiration came from studying human anatomy. In the human hand, the nail is not merely decorative. It increases local stiffness at the fingertip, limits soft-tissue deformation, and concentrates contact pressure, allowing fine-scale force to be applied with precision. Research on human subjects has found that people with approximately 2 mm fingernails show significantly enhanced manipulation dexterity. Despite this, artificial nails in prior robotic hand designs had been used mainly as simple grasping aids, without deeper examination of how they alter force transmission or contact mechanics.
The PLATO Hand — short for Proprioceptive Linkage-driven Anthropomorphic Hand — addresses this gap with a hybrid fingertip that pairs a rigid nail with a compliant, deformable pulp. The nail increases the fingertip’s resistance to bending, which redirects deformation inward toward the contact point rather than allowing the whole fingertip to buckle. The researchers showed that higher nail stiffness concentrates over 95 percent of deformation energy at the local contact zone rather than across the broader fingertip structure.
To test the design, the team mounted the three-fingered, eight-joint hand on a robotic arm and ran it through a series of demanding tasks. Without the fingernail, the robot failed at paper singulation, coin picking, card picking, and orange peeling in virtually every trial. With the fingernail present, success rates rose to 80 percent or higher across these same tasks. Pinching stability on flat and curved surfaces improved by 23 to 78 percent, depending on the surface geometry.
The researchers note in the paper that robotic hand design “must move beyond morphological mimicry toward functional principles that govern human fingertips, including how nails structure contact geometry and enable force-regulated interaction.”
The hand also incorporates quasi-direct drive actuators and a five-bar linkage mechanism designed to keep mechanical resistance low, enabling the system to sense contact forces through its own motors rather than relying entirely on dedicated sensors. Proprioceptive force estimates tracked directly measured forces with less than 5 percent error, even during rapid impacts.
The researchers write that “coupling structured contact geometry with a force-motion transparent mechanism provides a principled, physically embodied approach to precise manipulation.”
Journal Reference: ArXiv: 10.48550/arXiv.2602.05156
