Attrac Lab announced on August 25 that it has developed indoor drone-controlling technology utilizing AR markers in a joint project with Saga University.

The technology uses overhead cameras to recognize AR markers installed on drones and implement 3D-odometry. By being able to calculate location information across multiple cameras installed in different areas, testers were able to control drones through a wide range of spaces, such as long hallways, which cannot be covered by a single camera.

Attrac Lab

Images from cameras used to check drone's flight path that were attached to ceiling pointing toward the ground (top two) and images from cameras used to pick up AR markers attached to drone (bottom two).

The AR marker control algorithm was developed by Professor Kazuya Sato from the Faculty of Science and Engineering in the College of Natural Sciences at Saga University's Institute of Education and Research. Attrac Lab was responsible for developing the drone's autonomous control mechanics and transmissions.

Normally, controlling a several-centimeter large drone requires Lidar and a stereo camera. With this system, however, cheap web cameras can be used instead, giving the method notable advantages in terms of cost.

Moving forward, developers hope the technology can be applied to fields with large cost constraints such as agriculture, where it can be used for projects like controlling self-driving rovers (autonomous vehicle robots) inside greenhouses.