Drones with the ability to follow the user and shoot video or still images are out there right now. The catch is that they only really work in environments where there are no obstacles that the drone needs to avoid. Put the same tracking drone into a forest environment with random tree branches and trunks and the drone is unable to avoid those obstacles.
A company called Skydio has unveiled a drone that is capable of following a user and shooting video or still shots in a complex random environment like a forest path. The video shows the drone as it follows a person on a bike and dynamically avoids obstacles while keeping up with the rider. The video of the drone in action appears to be unedited, which is a good thing.
The wild camera array on top of the drone is where all the data the drone needs to avoid obstacles comes from. Skydio's Adam Bry said, "All the footage is just taken from our normal testing. Nothing was staged and it’s all fully autonomous. All of the navigation is done entirely based on a multi-camera array with all computation done onboard on a state of the art mobile CPU."
The two behind this slick drone are Bry and Abe Bachrach, the duo were formerly lab mates at Google X and had worked on Project Wing at MIT. The duo is using much of what was learned at MIT to create Skydio. The hardware in the Skydio drone promises to path-plan around obstacles. The done is a prototype, but the company has raised $25 million in funding to continue its work. This tech could one day lead to delivery drones able to get packages cheaply and safely to customers in real world environments.