Optimal altitude allocation for drone flights
The question of “Where do drones fly” is a fundamental question that we should be able to answer.
At present, a drone pilot makes an arbitrary decision to fly at a certain altitude, and most of the time an altitude ceiling is set by the drone manufacturer based on FAA/NAA safety regulations. Current U-Space/UTM providers such as AirMap, Unifly, Altitude Angel etc. only provide drone users with situational awareness with respect to active NOTAMs, No-Fly-Zones and, in rare occasions other unmanned traffic. Thus far, we have seen a significant gap in one fundamental aspect of future high-density operations ─ capacity management of the urban airspace. However, before we optimize capacity of the urban airspace, we need to ask ourselves “where do we assign drones to fly?” i.e., do they fly above or between buildings. This is an optimization problem and, it is imperative that we solve this problem first.
The envisioned framework should process the enlisted decision variables (not limited to the ones mentioned) by optimizing it with respect to a cost function. This cost function can be a something similar to the Cost Index (ratio between time cost and fuel cost) employed by airlines. The challenge here is that we deal with high densities of traffic, the optimization problem at hand becomes intractable. We assume that this can be solved via an AI method.
- A literature review
- Comparative industry analysis e.g. how airlines decide on routes
- Gather data on any existing U-Space/UTM services
- Formulation of objective function
- Decision variables
- Building a decision engine
- Low traffic densities
- High traffic densities
- Optimizing the decision engine in Blue Sky
The end goal of this assignment should be a framework/model that provides a data-driven answer to the question of “do drones fly above or between buildings?”.