Artificial Neural Networks-Based Robustness Prediction of Bilateral Teleoperated Scooping Motion

In this work, we have tried to predict the robustness of grip of a robotic arm with three finger end-effector by using the deep learning aspect of machine learning, while performing the gripping task with the use of bilateral teleoperation. Existing knowledge is identified by doing Literature review and grouping them together to carry the prediction task. The object of this research is to see how robustness is affected by changing different working parameters of the three finger end effector and it will help us to know the robustness of the grip of the three finger end effector for the new real life working values of the parameters of the three finger end effector. This work will show integration of deep learning with robotics and bilateral teleoperation and will help to find out working limitations between robotics and bilateral teleoperation. This work also shows why deep learning is chosen over machine learning to perform the prediction. In the end the paper will also discuss results and conclusions with future scope.