Air routing has become an important problem of recent years. Wide implementation of idea to use a free routing airspace (FRA) over the Europe and idea of exploiting FRA as a main airspace management resource to reduce air traffic problems revealed a necessity of a new look to a routing problem. Many previous solutions relied on predefined topology of airways and ability to exploit welldeveloped methods known in graph theory. Meanwhile the problem was current due to many factors needed to be involved in the airspace as a 3D-space: air management restrictions and different air spaces regulation rules, weather conditions, danger areas, aircraft’s characteristics, pilots’ preferences, etc. Moreover, the appearance of FRA has made it inappropriate to use previous algorithms. Most of these algorithms required a definite topology with known routing points connected with predefined edges, while the FRA may have only border points to fly into or fly out of the area and no definite edges inside. The task of constructing the route became the same difficult as obvious: any pilot can fly directly through the FRA, but the route should be built and confirmed prior to a take-off. Problem comes even more evident if considered for the unmanned flying vehicles (UFV) and the need for robots or AI systems to solve the routing problem by itselves. As a topping of the complexity of the problem, one may consider the upcoming difficulties of airspace congestion in FRA. Despite the problem is known for areas close to airports, it is still current to plan routes avoiding flights conflicts in the air and to avoid FRA high congestion. There are different researches on some particular problems and some approaches to solve these problems. Nevertheless, there is no complex problem statement yet. This research was focused on need of understanding the full scope of problems for air routing to understand the ability to build an efficient solution for the problem as a whole.