Scientists from the University of Zurich have created a new algorithm that controls the drone faster than a person.
A research team from the University of Zurich (UZH) has created an algorithm that can find the most optimal trajectory to fly a quad-propeller drone.
The algorithm-driven drone covered the experimental racing distance faster than world-class pilots. The authors believe this is not surprising because once the algorithm finds the best flight path, it can accurately reproduce it many times, unlike humans.
The novelty of the algorithm lies in the fact that it generates time-optimal trajectories that fully take into account the limitations of unmanned aerial vehicles. Previous work was based on simplifications of either the drone system or the flight path description, and therefore they were sub-optimal.
Davide Scaramuzza, Head of Robotics Group
Also on the drone are external cameras that provide real-time location information.
Now the authors of the new work are optimizing the new algorithm, since while the computer needs up to an hour to calculate the time-optimal trajectory for the drone.
In addition, at the moment, the drone relies on external cameras to calculate where it was at any given moment. Researchers want to use onboard cameras. Despite this, the authors see great potential in their work.