Mobile system with real time route learning using Hardware Artificial Neural Network

This article presents a solution for tracking a mobile system using an artificial neural network. The mobile system collects data from the environment using an ultrasonic transmitter and receiver then data is processing using a binary artificial neural network. Some templates have been pre-loaded into the system to avoid blockages or additional routes. The solution is implemented on a SOC manufactured by Xilinx: Zync7000 Artix which consists of an FPGA and an ARM processor. The FPGA contains hardware neural network, command units and acquisition unit, while the processor contains the interface that provides patterns for learning process and communication interface (Ethernet interface).