Design and Implementation of Self Learning Autonomous Robot using Neural Networks under ROS (Robot Operating System) Platform
Loading...
Authors
Al-Khammasi, Salam
Qayym, Adnan
Sobh, Tarek M.
Issue Date
2015-03-27
Type
Presentation
Language
en_US
Keywords
Robot , Arduino , Neural network , Ultra-sonic sensor
Alternative Title
Abstract
In this poster, we are designed and implemented an avoidance obstacle robot by using ROS (Robot operating system) as main platform. Neural Network algorithm has been used to program the robot. The algorithm has been written in Python programming language. In the hardware part, Arduino Uno Board with Ultra sonic sensor have been used to to detect the obstacles. Our contribution will be how to make the Robot detects the obstacle using neural network by learning himself from the environment and save that data which is getting by the ultra-sonic sensor to the base station so when it comes back to the same environment, the robot will not need to do the same procedure because the data already saved to the Base Station. All the related variables like Velocity, Acceleration and distance, etc. will get from ROS platform. The ROS will minimize the coding and gives us relative results. The communication between the robot and the base station will be wireless.