| The new tendency in mobile robots is to create non-Cartesian systems based on reactions to their environment. This emerging technology is known as Evolutionary Robotics, which is combined with the Biorobotic field. This new approach brings cost-effective solutions, flexibility, robustness, and dynamism into the design of mobile robots. It also provides fast reactions to the sensory inputs, and new interpretations of the environment or surroundings of the mobile robot. The Subsumption Architecture (SA) and the action selection dynamics developed by Brooks and Maes, respectively, have successfully obtained autonomous mobile robots initiating this new trend of the Evolutionary Robotics. Their design keeps the mobile robot control simple.; This dissertation work presents a biologically inspired modification of these schemes. The hippocampal-CA3(a biological type of neuron)-based neural network (HCA3) developed by Williams Levy is used to implement the SA, rather than augmented finite state machines (AFSM). The HCA3 is based on the hippocampal cortex region of white rats. This replacement by the HCA3 results in a closer biological model than the SA, combining the behavior-based intelligence theory with neuroscience. The design is kept simple, and is implemented with the Khepera Miniature Mobile Robot. The HCA3 handles complicated situations beyond its expected limits. This simple system behaves like a complex one, a phenomenon observed in nature where computational speed is not as critical as in robotic or computer applications. The action selection dynamics emerge from the iterations of the levels of competence implemented with the HCA3. The resulted-autonomous mobile robot determines its action and the state of operation. Finally, the design is used to execute a mail delivery and surveillance robot inside a building floor.; The use of the HCA3 in mobile robots is its first application. The application adds practical aspects to the HCA3 that help to understand the network in more details. New techniques for tuning the network for fast and better performance have emerged from these practical issues. The gained insight increases the possibilities for future neural network developments. |
