Exploring SBC with Kinect And machine Learning for Indoor Mobile Robot Design and Monitoring

The aim of this paper is to present our way to develop and implement indoor mobile robot by embedding new hardware vision and computing device with ML models, exploring the full power of a Kinect Depth Camera using Deep learning algorithms and using the power of Raspberry Pi 5 to design a Mobile Robot that can follow a Human Companion indoor using Recognition of His facial ID then follow the orders given, orders can be given using had gesture, Body gesture ‘skeleton’ and voice commands. ML models have been explored and improved to implement this design. This is part of a project implementing DL and ML deep learning algorithms to achieve various applications. That include Face detection and recognition which then marks a specific person as the targeted companion, the robot then follows and reacts at his hand and body gestures in real-time. In the project, the prototype hardware design, the choice of components and embedding the right software and ML models. The objectives on this work are to improve the reaction and comportment of the mobile robot giving it more power computing and dept. vision like human and implementing and running multi ML models on it, the first experiments were interesting, we simulated the design and got good results.