Would it not be cheaper and faster to simulate a drone and fly it through virtual 3d environments and still learn?
Or would the physics be too complex to model well for simulation?
Or would the physics be too complex to model well for simulation?
The elegant thing about using machine learning is that you don't need to build any models at all. And you can develop the ML technique once and then reuse it to train different hardware configurations, instead of incurring the cost of modeling every one.
Also, note that the physics of the simulation doesn't even need to be realistic. Unless you are doing high-speed control or aggressive maneuvers, the challenging part is the perception and not the control. In the paper from OP, the controls are even high-level discrete actions: left, forward, right.