Robotic arm manipulated by eye movement using a webcam and computer interface can help people with severe speech and motor impairment

People with Severe Speech and Motor Impairment (SSMI) – a condition caused by disorders such as cerebral palsy – find it difficult to physically operate devices such as a joystick, mouse or trackball, or use speech recognition systems.

To help people with SSMI, a research team at the Centre for Product Design and Manufacturing (CPDM) has designed a robotic arm that can be manipulated by eye movement using a computer interface. This interface is non-invasive since it is through a webcam and a computer, unlike other eye gaze-tracking devices that use head-mounted systems.

The team worked with young adult students who have SSMI at Vidya Sagar (formerly The Spastics Society of India) in Chennai, an institution for individuals with disabilities. “Many of these students [with cerebral palsy] are not able to precisely focus at a single point in their visual field, due to uncontrolled gaze movement. They are also not comfortable to look at all portions of the visual field equally,” says project lead Pradipta Biswas, Assistant Professor at CPDM.

Biswas and his team used computer vision and machine learning algorithms to analyse live feeds of facial video from the users, and were able to estimate where they were looking. The team coupled this with an Augmented Reality application to allow users to use a robotic arm for tasks like picking up and dropping objects, and placing them where they desire, as long as it was within reach of the robotic arm.

One of the main applications of this eye gaze-controlled robotic arm is to rehabilitate young adults with SSMI through tasks such as fabric printing. They usually require assistance in doing such tasks, as they can only do a small part of it by themselves. Using the robotic arm designed by the researchers, people with SSMI can use their eye gaze to perform mechanical tasks that can help them work on handicrafts independently.

This eye gaze-controlled interface has been deployed at Vidya Sagar and is in regular use. “This validation and evaluation [of the interface and robotic arm] with end users is a major contribution of the study,” says Biswas.

The students with SSMI are able to use the robotic manipulator almost as well as their non-disabled counterparts.
This technology can also be used by younger individuals with SSMI to move toys like cars. “We are using play as a medium to teach new technologies, which they can use for the rest of their lives,” adds Biswas. According to the researchers, the tool could also be adapted for e-learning.

The authors believe that this interface is an important step towards designing technologies that would help those with physical impairments resulting from SSMI learn better and become skilled professionals. Biswas also says that such a system could be useful for automotive and aeronautical applications, as well as for developing collaborative robots used in smart manufacturing.