Wireless EMG-Driven Realtime Robotic Hand Control
Disciplines
Biomedical | Electrical and Computer Engineering
Abstract (300 words maximum)
There are currently thousands of amputees and people who have lost full use of their limbs across America. More often than not, these people have very low income or are in a situation where they cant take extended time off work. Prosthetic technology as we know it now has several limitations that heavily impact people like this the most. Often prosthetics that restore function use of a limb are very cost, requiring thousands of dollars for prosthetics with simple sensors or functions. These surgeries are also often very intrusive and come with lengthy surgeries and recovery times that impede every day life. This study aims to improve current prosthetic norms and technology by using Electromyography (EMG) sensors in limb prosthetics. EMG sensors detect changes in electricity flowing through muscles and using those changes, we hope to see if it can accurately be used to move limbs like hand or feet that a patient may have lost. EMG sensors are completely unintrusive and work through skin contact, unlike may similar sensors that require access to nerves. EMG sensors are also very cost effective and require little setup to be used. This study will test the feasibility of using EMG sensors as an accurate tool for prosthetics by seeing how well 8 EMG sensors connected to the forearm are able to control a robotic hand remotely. A data glove that measures changes in angles when the fingers are bent will be used to train an AI model that receives live EMG data. This model will then be used to predict the angles of each finger and correctly move the robotic hand accordingly.
Academic department under which the project should be listed
SPCEET - Electrical and Computer Engineering
Primary Investigator (PI) Name
Coskun Tekes
Wireless EMG-Driven Realtime Robotic Hand Control
There are currently thousands of amputees and people who have lost full use of their limbs across America. More often than not, these people have very low income or are in a situation where they cant take extended time off work. Prosthetic technology as we know it now has several limitations that heavily impact people like this the most. Often prosthetics that restore function use of a limb are very cost, requiring thousands of dollars for prosthetics with simple sensors or functions. These surgeries are also often very intrusive and come with lengthy surgeries and recovery times that impede every day life. This study aims to improve current prosthetic norms and technology by using Electromyography (EMG) sensors in limb prosthetics. EMG sensors detect changes in electricity flowing through muscles and using those changes, we hope to see if it can accurately be used to move limbs like hand or feet that a patient may have lost. EMG sensors are completely unintrusive and work through skin contact, unlike may similar sensors that require access to nerves. EMG sensors are also very cost effective and require little setup to be used. This study will test the feasibility of using EMG sensors as an accurate tool for prosthetics by seeing how well 8 EMG sensors connected to the forearm are able to control a robotic hand remotely. A data glove that measures changes in angles when the fingers are bent will be used to train an AI model that receives live EMG data. This model will then be used to predict the angles of each finger and correctly move the robotic hand accordingly.