Artial's software solution aims to enable
drones to operate autonomously. The company develops algorithms based on deep learning that process input data on board of the device. During the flight, the drone will not use a standard terrain map, but build a 3D-map in real time using computer vision. Thanks to this technology, the device avoids collision with static and dynamic objects, be it buildings or power lines, as well as "unexpected" objects that arise during flight, such as a bird. Smart drones will “understand” when to land, start moving, safely complete a maneuver or return back to the origin.
The advantage of the technology is the use of aircraft beyond the line of sight of a person, in dangerous places and areas of unstable radio signal. If a device with autonomous technology loses contact with the pilot, it does not fall, but continues to perform operations, but on its own.
The scope of drones with Artial filling is extremely wide: energy infrastructure, telecommunications, transport, including railway communication. All areas, says Egor, where constant checks, prevention of breakdowns and predictive maintenance are required.
For example, now the service team arrives at a specific site of the power grid, where there is a malfunction or a check is required. Workers manually launch the drone and pilot it. However, they are limited in operation within a radius of several meters. The closest alternative to avoiding the human factor is an inspection done by helicopters, but an hour of its flight costs between 6,000 and 8,000 euros. Therefore, Artial with autonomous technology sees its competitive advantage in expanding the scale of operations and significant savings.