Robotic framework for developing of unmanned vehicles

Abstract

Unmanned Vehicles, known as UVs, have been developed to accomplish difficult, tedious, and unpleasant missions for human beings. Their usage started in military applications; subsequently, specific industries adopt them to increase the productive capacity of their factories already automated with industrial robots relying on the mobility and autonomy offered by these vehi cles. However, unmanned vehicles could operate in many other sectors - like in agriculture, construction, logistic, customer service -, facilitating and im proving the quality of life in general. It is necessary to increase its develop ment and implementation substantially to achieve this. As we will show in this document, progress in the area of mobile robotics, especially in the field of unmanned vehicles, has been essential and prolifer ates. However, it is not robust yet to be reliable and accepted beyond the con trolled environments in which they operate nowadays. The enlarged area to cover, due to mobile capabilities, plus the risky missions they need to accom plish, increases the complexity of autonomous steering and control of such a vehicle. For this reason, the modeling considerations to achieve the task of autonomous driving is considered complex and reserved to humans (Litman T., 2018), due to the high frequency of interactions with other mobile objects, which requires sensing and acting capabilities in real-time bases, with an es sential degree of intelligence and skills. An essential step in their development is the robotics environments for de veloping and testing unmanned vehicles, these computational environments incorporate and centralizes all technologies, in their broadest sense, related to robotics. In both professional and academic literature, these environments are called with different names, such as robotic middleware, robotic platform, and robotic framework. Their degree of development and their capacities are not homogeneous, being those specialized and commercial branded who have reached essential levels of maturity and acceptance. As it is the case of X Plane, a platform for the simulation of autonomous flights of many well known aircraft, the Federal Aviation Administration (FAA) can even certify the implementation if they also count on certified hardware. Some conventional robot models are offered in these robotic environments, which are highly used for research in robotics. It is an essential contribution to the robotics community to get the research efforts to concentrate on the central issues of their work. However, it does not help much if there is a need to test a new robot model from scratch, where the initial main effort is in the modeling and evaluation of behavior in order to redesign the model itself. These environments also offer standard robotic functionalities that come to help in both cases. ... [edited by Author]