Robot taught to read braille twice as fast as humans


Researchers train a robotic sensor to read braille at 315 words per minute – roughly double the speed of average human readers.

The robot, developed by a research team at Cambridge University, uses machine learning algorithms to quickly scan the braille text. It deciphers words at a rate of almost 90% accuracy and uses a camera as its “fingertip,” according to the university.

Human fingertips are remarkably sensitive, which is one of the reasons why we are able to grip things with the right amount of pressure and detect tiny changes in the texture of a material, it said. Recreating this sensitivity in a robot is a big engineering challenge.

“For robotics, softness is a useful characteristic, but you also need lots of sensor information, and it’s tricky to have both at once, especially when dealing with flexible or deformable surfaces,” said the study’s first author, Parth Potdar from Cambridge’s Department of Engineering.

In addition, images captured by a camera need to be processed to remove the motion blur, which consumes time and energy.

To overcome this challenge, researchers trained the algorithm on a set of sharp images of braille with fake blur applied. This allowed the robotic reader to “deblur” the images before the sensor attempted to recognize the letters.

“Considering that we used fake blur to train the algorithm, it was surprising how accurate it was at reading braille,” said co-author David Hardman, also from the Department of Engineering at Cambridge University.

Hardman said it was a “nice trade-off” between speed and accuracy, which is also the case with human readers.

The robot was not developed as an assistive technology, but researchers said that the high sensitivity required to read braille made it an ideal test in the development of robot hands or prosthetics with comparable sensitivity to human fingertips.

They said the robot could also be used for applications like detecting surface textures or slippage in robotic manipulation and highlighted plans to scale the technology to the size of a humanoid hand or skin.

The findings of the study were published in a paper entitled “High-Speed Tactile Braille Reading via Biomimetic Sliding Interactions” and published in the journal IEEE Robotics and Automation Letters.


More from Cybernews:

Harvard scientists build a robot gripper with spaghetti-like tentacles

Deceased spider inspires new field of "necrobotics"

Google DeepMind presents self-training RoboCat

Google crafts helper robot to make you snacks and wipe your table

Researchers design remotely-controlled "biobots" powered by mouse cells

Subscribe to our newsletter

Subscribe to our newsletter



Leave a Reply

Your email address will not be published. Required fields are markedmarked