Cambridge researchers have crafted a robotic sensor using AI to read braille twice as fast as humans. The machine, trained with machine learning algorithms, achieved 315 words per minute with 87% accuracy. While not designed as assistive tech, its sensitivity makes it valuable for developing robot hands or prosthetics with human-like touch. Professor Fumiya Iida’s lab tackles the challenge of replicating human fingertip sensitivity in energy-efficient robotic hands. The team’s innovative approach involves a camera-equipped ‘fingertip’ using machine learning algorithms to enhance image clarity, paving the way for broader applications beyond braille.