Immunity-Based Dynamic Reconfiguration of Mobile Robots in Unstructured Environments

Mobile robots have shown the capability to address difficult challenges ranging from disaster response to autonomous delivery of goods. However, the robots designed for one environment are often unable to perform effectively in other settings, and frequently require reconfiguration on software and/or hardware levels. Therefore, authors have designed a new immunity-based arbitration scheme to dynamically reconfigure mobile robots during their activity in unknown environments. The technique continuously selects the most suitable modules from its library e.g. sensor-configuration and navigation approach, on the basis of immunological mechanism of clonal selection. The technique successfully arbitrates different robot-configurations because of its ability to clone, mutate and select the suitable antibodies as well as to dynamically change the antibody concentrations. Moreover, maintenance of low inflammation levels alongside low collision count and slow battery drainage clearly illustrates the effectiveness of the presented technique.