Using machine learning to improve prosthetic limbs

Using machine learning to improve prosthetic limbs

Researchers from the University of Erlangen-Nuremberg are investigating how intelligent prosthetic limbs can be made more reliable using machine learning.  

Researchers are trying to find new ways to make interactive artificial intelligence (AI) prosthetic limbs better recognise human intent. The team are developing AI that can register their surroundings and improve functionality over time.  

Robotic prosthetic limbs already allow people to perform important everyday functions such as lifting items, making a fist and entering phone numbers. However, current models often encounter stumbling blocks with regular use. Many of the current prosthetic limbs offer a ‘one size fits all’ model that fails to accommodate the diversity of each individuals’ intentions.  

Improving the limitations of prosthetic limbs

“We are literally working at the interface between humans and machines. The technology behind prosthetics for upper limbs has come on in leaps and bounds over the past decades,” explained Dr Claudio Castellini, professor of medical robotics at the University of Erlangen-Nuremberg. 

Dr Castellini heads the Assistive Intelligent Robotics lab at the university. The lab focuses on creating assistive robotics for the upper and lower limbs as well as functional electrostimulation. Dr Castellini’s team focuses specifically on the bio signal processing of sensor modalities and methods of machine learning for intent detection.  

By using surface electromyography, skin electrodes at the remaining end of the limb can detect the slightest muscle movements. These bio signals are then converted and transferred to the prosthetic limb as electrical impulses. 

“The wearer controls their artificial hand themselves using the stump. Methods taken from pattern recognition and interactive machine learning also allow people to teach their prosthetic their own individual needs when making a gesture or a movement,” said Dr Castellini.  

Many people still prefer non-robotic limbs 

Advanced robotic prosthetics have not yet reached their full potential in terms of comfort, function and control. For this reason, many people with missing limbs still prefer to use purely cosmetic prosthetics with no additional functions. 

A new EU Horizon project, ‘AI-Powered Manipulation System for Advanced Robotic Service, Manufacturing and Prosthetics’, is concentrating on how new prosthetic limbs can interact with their environment more effectively and complete more specific functions. 

Researchers at the University of Erlangen-Nuremberg are focusing particularly on how to improve the control of both real and virtual upper limb prosthetics, this is known as intent detection. Dr Castellini and the team will be analysing and recording human bio signals to help them design innovative algorithms for machine learning aimed at detecting the individual movement patterns of individuals. 

The researchers studied people both with and without physical disabilities in order validate their results. Furthermore, the university is leading “shared autonomy between humans and robots” section of the EU project, aimed at checking the safety of results. 

“We are exploiting the potential offered by intent detection to control assistive and rehabilitative robotics. This covers wearable robots worn on the body such as prosthetics and exoskeletons, but also robot arms and simulations using virtual reality,” said Dr Castellini. 


Please enter your comment!
Please enter your name here