In this video you can see a prototype of our experimental drone set up with a companion computer and an advanced lidar system during the checks of the anti-collision system that working simultaneously with the localization and mapping algorithms allows the drone to avoid obstacles during autonomous flight .In the last 5 years in the Laser Lab, localization, mapping and anti-collision algorithms have been developed, using 2D and 3D type lidar.In this first part of the post we will focus on algorithms using a 2D lidar and a 2D SLAM navigation algorithm.
The videos of the playlist are made using:
The outputs of the navigation system are:
In the following video we tested the SLAM algorithms in the corridors and the offices of our laboratories , the quality , accuracy and repetition of the location was reliable and allows to implement excellent autonomous navigation algorithms without GPS and with anti collision, as we will show in the following examples.
After the excellent results obtained in the ground tests we carried out the tests in flight by installing the localization system on board of our drone injecting the location data instead of the signal normally sent to the VR Brain 5.2 from the GPS, in addition to the position signal provided from the on-board computer we have also interposed the heading direction data generated by the navigation algorithm to the autopilot in place of the data provided by the magnetometer. The magnetometer has been replaced because in indoor areas the perturbation of the magnetic field is such as to compromise the navigation system making it unstable.
In the video you can appreciate the quality and accuracy of the location, the accuracy of the heading direction and the low latency of the calculation algorithms for localization and mapping.
The height data is given by the barometer and recurrent by the on-board accelerometers of the VR Brain 5.2.
In the next video we see an example of the flight telemetry of an autonomous mission inside a boiler, the environment is a metallic environment where the position could not be provided through a traditional GPS, the only possible option it was set up a drone able to navigate autonomously thanks to the help of a system of localization and mapping based on a laser system.
Now a couple of examples of application , the first flight was a video inspection on a vertical wall , the flight planning did on the ground station sent to the drone and the flight did independently without pilot.
On the next flight, instead, an example of a horizontal inspection flight.