Gesture control robotic arm using Myo armband

Project Title

Gesture control robotic arm using Myo armband

Project Area of Specialization RoboticsProject Summary

Human gesture interaction with robotics system has become an important in control system because it provides more precise, efficient and sustainability control. This project discuss about the design approach of robotic system control by using muscle sensor .The muscle sensor that has been used for this project is a MYO Armband sensor which identified the electromyography EMG signals based on the contraction and relaxation of muscles activity .when human perform any gesture this project aims to first train the MYO armband to learn gesture through artificial neutral networks and then control the robotic hand through the MYO amrband the outcome of this project is the robotic hand that successfully performs a gesture motion according to intuitive interaction of human

Project Objectives

1.To connect with MATLAB.

2.To collect the raw EMG data from MYO armband which are 5 gestures

3.Then train the row EMG data through artificial neural networks to recognise the gestures finally these gestures are generated by robotic hand which uses the servo motor mechanism for motion of hand.

Project Implementation Method

First of all we connect our Myo armband with MATLAB we will extract EMG data on MATLAB we will start the training of Myo armband using artificial neural network that networks where created the in MATLAB using NPR tool interface using MYO armband a large set of data was recorded As the user perform each gesture 5 times it is input of the neural network then we sat target values corresponding to the input each row indicates one type of gesture the recorded training data corresponding just last data are fed to the network training tool and start training the created network consists to 5 input nodes hidden layer consisting of 10 nodes and output layer consisting of 5 nodes one for each gesture the number of input and output nodes naturally correspond to the number of inputs and outputs used while the number of nodes in the hidden layer can be customised 10 was found to be the suitable number for reading more notes did not cause any significant improvement in performance using the data using the training data network on so many titrations just the ways between the nodes and improving the classification of the training set after training a robotic hand recognise this the five gestures now MYO armband is connected with Ardiuno through laptop Myodiuno software wirelessly and our robotic hand also connected to our laptop through wires and we add sketch of robotic hand control in ardiuno now we perform gesture and robotic hand perform the gesture that we program in ardiuno Myo armband send the E M G data to laptop and laptop send it to arduino and accordingly.

Benefits of the Project

The robotic hand controlled by MYO armband can be potentially used for the disabled person who has lost his hand to accident and it is also as surgical device the surgeon will be able to control the movement using MYO armband and the robotic hand performs a surgery on the patient precisely the robotic hand replicate every moment that surgeon make robotic hand can be built in microscopic size where it can easily fit through very small incision which can potentially minimise scarring of the patient further more it can provide the surgeon with unprecedented control in a minimally invasive environment and is more hygienic and the reduce the chance of post operative infection additionally the robotic hand control can be found useful in environment where it is dangerous for humans in space or place such as radioactive environment where the robotic hand would be useful or it could be used to control and check bombs it distance

Technical Details of Final Deliverable

This is a highly multidisciplinary project combining electrical, electronics ,and manufacturing, computer and even chemical knowledge therefore the project has been marked by learning a continuous improvement trying to constantly overcome the challenges that each part of the project has been presenting in this way the learning of MYO armband is increased through ANN artificial Neurol networks and successfully controlled robotic hand through myo armband with small expectation the reverted hand can only be controlled by custom myo armband gestures according to you desire you can also modify the robotic arm by giving them a wrist rotational movement decrease current weight and size or continue to develop the rest of the arm even build a complete robot potential application is the handling of dangerous substances because thanks to robotic hand any kind of human contact with these substances avoided beside allowing to interact with then total security provides great precision and realism because it is not being controlled by computer it is controlled by operator on hand thanks MYO armband.

Final Deliverable of the Project Hardware SystemCore Industry ManufacturingOther Industries Others Core Technology RoboticsOther Technologies Wearables and ImplantablesSustainable Development Goals Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	65600
Ardiuno	Equipment	1	1850	1850
servo moter	Equipment	5	2650	13250
Myo armband	Equipment	1	45500	45500
3d printed robotic hand	Equipment	1	5000	5000