Field interactive robotic entity (FIRE) adaptive learning of environment using neural networks a VLSI implementation

Field interactive robotic entity (FIRE) is an ATMEL/spl trade/ micro controller controlled (belong to the 8051 microcontroller) robot. The maneuvering capabilities of the robot are based on two sensors, touch and edge detector. The touch sensor is a switch, to produce a low signal when an obstacle is encountered. The limitation of the robot is that it can sense an obstacle only when it makes a contact with it. To overcome this limitation an application specific integrated circuit [ASIC] chip is proposed incorporating neural networks. The learning algorithm implemented is 'unsupervised learning' based on memory. The proposed design is aimed at implementing in an XILINX/spl trade/ Spartan FPGA.