Developers: | Cornell University |
Date of the premiere of the system: | May 2024 |
Branches: | Pharmaceuticals, Medicine, Healthcare |
2024: Product Announcement
Researchers at Cornell University have developed a robotic feeding system for people with limited mobility. This was reported on May 9, 2024 by TechXplore.
The development uses computer vision, machine learning, and multimodal sensors to feed people with severe mobility limitations, including those with spinal injuries, cerebral palsy, or sclerosis.
According to one of the authors of the development, the training of robots in the feeding process took years. This is a multilevel process, starting with the robot identifying what is food on the table, before lifting the food and feeding the person neatly.
The last five centimeters, from a dining appliance to a person's mouth, were the most difficult - said Tapomayu Bhattcharji, one of the main authors of the development. |
Difficulties are also due to the fact that some patients may have limited mouth movement (up to less than 2 cm), and some patients suffer from involuntary muscle contractions, which can occur even with a spoon in the mouth. Many patients can only bite food with a specific part of their mouth that they reach by moving the spoon with their tongue.
The available technologies look only at the face of a person and assume that it will be motionless, although this is often not the case, said Rajat Kumar Jenamani, the main author of a scientific article on the development. |
To solve the existing problems, scientists equipped the robot with a real-time mouth tracking system that adapts to human movements, as well as a dynamic response mechanism that allows the robot to calculate the nature of physical interaction and act on this. So the robot is able to distinguish muscle spasm from a food bite or an attempt to move the device into a convenient part of the mouth.[1]