Researchers at the University of Cambridge have created a robotic sensor that can read braille twice as fast as humans. The robotic sensor glides over braille text using advanced machine learning algorithms, reaching reading speeds of around 315 words per minute with nearly 90% accuracy.
What does this mean?
The performance represents a big step forward in touch sense and robotic dexterity.
While the robot’s initial development goal was not to assist visually impaired people in reading, its effectiveness in high-speed braille reading demonstrates the possibility for improvements in robotic sensitivity.
Braille, with its emphasis on increased tactile sensitivity, offers an excellent test case for the creation of robotic hands or prosthetics that can mimic the sensitivity of human fingers.
This research was spearheaded by Professor Fumiya Iida’s lab at Cambridge’s Department of Engineering.
The study’s original author, Parth Potdar, emphasized the relevance of softness in human fingertips for holding objects with the appropriate amount of pressure.
However, recreating this quality in a robotic hand is a considerable engineering problem, especially when dealing with flexible or deformable surfaces.
Why was braille used as a testing ground?
The researchers chose braille as a testing ground because it requires a high level of tactile sensitivity.
Unlike standard robotic braille readers, which operate letter by letter, the Cambridge team created a more efficient and realistic reading technique.
Their robotic sensor, which has a camera in its “fingertip,” uses visual and sensory input to interpret braille.
The scientists also trained their machine learning algorithm on blurred Braille images, which improved the system’s character recognition.
This advancement extends beyond braille reading and has larger implications for robotic applications that require tactile sensing, such as texture detection or slip prevention during object manipulation.
The findings were reported in the journal IEEE Robotics and Automation Letters.
