The Mavic combines dual forward and downward vision sensors to realize precision hovering. Therefore, it requires no more than a patterned surface and enough available light to be able to hover stably up to 13m above the ground. Even if the Mavic is unable to identify reference points on the ground, it can still rely on its forward vision sensors to hover. In contrast, an Optical Flow system, requires information from ultrasonic sensors to supplement its view of patterned surfaces.
An example to show the difference between the two systems is when the aircraft is returning to you where you are standing on a balcony. When flying in over the balcony, drones using Optical Flow will receive data from two data sources that measure its height above the below surface differently – one being its height above the ground, and another the height above the balcony. As the data is not compatible, an Optical Flow system may become confused.
Another example is when the ground below the drone is covered by vegetation. Plants and trees do not reflect sound as well as other surfaces, causing the Optical Flow system’s ultrasonic sensors to be unable to gather altitude data.
In conclusion, the Mavic’s Intelligent Vision Positioning System requires fewer inputs and is therefore more robust in its functionality.