IIT-K Professor Behind Chandrayaan-2 Now Build Robot Exoskeleton Hand to Help People with Paralysis

KANPUR: Taking calibrated steps inside the Centre for mechatronics at IIT-Kanpur, professors Ashish Dutta (mechanical) and K S Venkatesh can hardly contain their excitement on the Chandrayaan-2 launch mission, which lifted off on July 22.

The IIT-K duo designed two subsystems for Pragyan — the six-wheeled rover that the Indian Space Research Organisation (ISRO) sent aboard the GSLV-MKIII by this week. The subsystems — light-based map generation and motion planning — are crucial for the success of the mission.

But, the professors also have another reason for their delight — they recently completed trials of the world’s first robot exoskeleton hand that will significantly help in the rehabilitation of stroke patients.

“It’s a proud moment and double joy for us, while we recall the algorithm, which we made for lunar mission Chandrayaan to help the rover move safely between one point and another, known as path planning or motion. The structured light is like a scanning laser that will survey the surface of the moon, take pictures and generate 3-D maps and path planning. This will ensure four aspects — take minimum energy (20-50 watts) on 10 MHz processor, safe route, prevent rover from toppling over and avoid obstacles in its route. Now, our next mission is completing large-scale clinical trials of the exoskeleton in India and the UK,” said Dutta, who is coordinator of the Centre for mechantronics.

The two-finger device is worn by a patient on his hand. It interprets brain signals, with the help of a brain computer interface (BCI) worn on the head, and helps paralytic patients open and close the motion of his thumb and index and middle fingers for physical practice. The robotic arm uses four bar mechanism and has four degrees of freedom (DOF). It’s driven by an AtMEGA 300 Mhz micro-controller and powered by a battery.
For the project, the duo teamed up with UK-based Ulster University and its professor Girijesh Prasad, who comes from Gorakhpur. Trials took place over six-weeks on four hemiparetic stroke patients (screened out of 16) with left hand disability in the UK and later on a patient in Kanpur. The professors say the device will cost around Rs 15,000.

HOW THE ROBOTIC ARM WORKS

EEG (electroencephalogram) electrode is used to acquire brain signals using a brain computer interface (BCI) worn on the head

Signal transferred to robotic hand — a two-finger device — known as exoskeleton

Sensors used to monitor patient’s finger-tip pressure

Anthropomorphic visual feedback received on computer

A year ago, Indian-origin scientist Ravinder Dahiya, professor of Electronics and Nanoengineering at University of Glasgow’s School of Engineering, had created a robotic hand called “brainy skin ” which recreates the human touch. It reacts like human skin, which has its own neurons that respond immediately to touch rather than having to relay the whole message to the brain.

Dutta and Venkatesh’s device works through sensors, which monitor the finger-tip pressure applied by the patient. If the patient is able to close or move his finger, the device passively follows the motion. If he does not, then device actively forces the finger to close, while taking directions from BCI using EEG signals. As far as design of the device is concerned, the exoskeleton’s degrees of freedom movement is based on the human finger motion while manipulating a coin in the hands. The joints on the device consist of four bars to give a human motion.

“We were glad with successful results. Next stage would that be of using exoskeleton for not just physical practice but for routine home work,” said Prasad. The Rs 55-lakh MHRD project was sanctioned in 2018 and led by its department of science and technology and British Council in the UK. The research has already been published in leading journals including Journal of Neuroscience methods; Biomedical and health informatics; Haptics and Engineering in medical and biology society.

---

Source: ToI

Image Courtesy:NewsJizz

---