Configuration of an Autonomous Vehicle a Systematic Framework
Gutiérrez Jagüey, Joaquín
MetadataShow full item record
Autonomous Vehicles have generated an emerging interest in demanding applications due to their great potential to move toward a particular site and perform some specialized work (such as goods transportation, exploration, and work tool manipulation), especially in remote environments which can be dangerous, unsuitable, and inaccessible for humans. Despite the great potential of Autonomous Vehicles, including safety and productivity, their use is not widespread because of the complexity to define the robotic vehicle configuration. Besides, current design processes rely on the designer´s experience, require long periods of time and signify high investments. Thus, practical approaches for their configuration are needed. This thesis focuses on the development of a method for the synthesis of an Autonomous Vehicle configuration, based on the analysis of both the task and the environment. The method assists to determine a suitable configuration to that serves as a comprehensive foundation in the building of the robotic vehicle. The configuration is expressed as a hierarchical and modular structure of interlinked components that fulfill the requirements to perform a given task under the constraints of a certain environment. The components implement the functionality of a composition of motions that solve the task. This composition is selected through a preliminary analysis, in which a geometric description of the task and the environment is used to combine motions and to find viable compositions (possible solutions). Each motion of the selected composition becomes a component or component set into the vehicle structure. To select the components, the method uses criteria of Robotics and engineering principles for providing of autonomy to the configuration, as a function that correlates the perception, the control, the positioning, and the geometry of the robotic vehicle. The definition of the requirements, the geometric description, the combination of motions, the selection of the fundamental composition, and the gradual fulfillment of the configuration with the proper components are derived through this systematic framework, exploiting the vast inventory of technologies and products developed by decades of engineering and robotic research. The implementation of this method is illustrated through a real request for the underground mining domain. The major result of this research work has been the formulation and validation of the proposed framework as a suitable approach to systematically determine the Autonomous Vehicle configuration.