Artificial Intelligence-Based Human-Assisted Multipurpose Robot

This paper explains the most important methods for self-learning systems to develop a self-learning algorithm and a suitable demonstrator for giving an insight in applications and state of the art in self-learning systems. The system described in this task is realized with Q-learning as table method. The agent perceives the environment employing sensors mounted in front. The simulator has been developed in MATLAB to be as close as possible to the physical model. The physical model can either be trained up by reason or benefit itself of finished simulated values. Simulation results show that the table method makes the simulated agent to navigate in an unknown environment, while the physical model only handles static obstacles due to physical limitations and project time scale. An artificial intelligence-based working prototype has been designed to control a robot wirelessly using an Android application. In completing this prototype, wireless software and hardware technologies have been used, such as NodeMCU—ESP8266 wireless module, a dual-channel H-bridge L298N IC for motor driver module, and four electric DC motors connected in parallel are used to move the automobile. The Android application is used by the user to send data wirelessly by connecting to the NodeMCU module using WiFi. This data is an input to the microcontroller system and the microcontroller uses it as the controlling parameter to the underlying hardware. The advantage of using this robot is that it can be used for multiple purposes which include a spy camera as well that can stream the videos to the user over Wi-Fi.