A key challenge in most UAV-based applications is to realize a safe UAV flight planning and navigation. To achieve the highest possible level of safety, risks to other airspace users and to people and property on the ground must be minimized. Considering current UAV regulations, conducting a safe flight mission relies on accurate and up-to-date geospatial data, as it serves as a basis for UAV operations and therefore plays an important role in UAV mission planning. Consequently, UAV path planners (in most cases) use maps like Open Street Map to obtain the desired data such as waypoint coordinates. Here, the question is to which extent these maps represent the UAV environment and the available data are sufficient to plan a flight path in the best possible way so that the subsequent flight mission can be carried out successfully. Consequently, an adequate representation of the UAV environment should not only focus on spatial aspects like the geometric representation of objects, but also on the so-called semantics. Semantic data describes the object structure of an environment, including contextual information, attributes and their relationships among each other such as building type, usage, address, etc. However, in UAV planning, a considerable amount of semantic information is currently not considered; and this may lead to a lack of information that may affect the interaction between UAV and its environment and consequently the flight path may not be generated optimally. From this point of view, we aim in this research area to develop a prototype which enables safe UAV flight planning considering current UAV rules. This prototype should generate a hybrid Dynamic 3D Map (HDM) for safe UAV planning in the form of a link between current geometry of the UAV environment and current contextual and semantic information. The workflow of creating the mentioned HDM is shown in the Figure below.
The goal is to develop a prototype which enables safe UAV flight planning considering current UAV rules. This prototype should generate a hybrid map for safe UAV planning in the form of a link between current geometry of the UAV environment and current contextual and semantic information. The goal can be achieved by three steps:
