Top.Mail.Ru
  • SSUGT
  • SSUGT news
  • SSUGT students developed software for emergency stopping of a multi-rotor drone when an obstacle is detected

SSUGT students developed software for emergency stopping of a multi-rotor drone when an obstacle is detected

The Siberian State University of Geosystems and Technologies tested the software for emergency stopping of a multi-rotor drone when an obstacle is detected, developed by the university's youth engineering team at the request of Mirai Vision LLC.

Recall that in July 2024, following an open selection of providers for training participants in youth engineering teams as part of building flexible educational trajectories of the Federal Project "Personnel for Unmanned Aircraft Systems", conducted by the ANO "University of the National Technological Initiative 2035", SSUGiT received the status of a Support Center.

According to the selection conditions, SSUGiT had to develop an additional education program and prepare a youth engineering team to solve specific problems in the field of maintenance and design of unmanned aircraft systems.

As part of the project, SSUGiT scientists developed an educational program "Improving the safety of using UAVs: developing software for UAVs equipped with a single-board computer". The program provides the necessary competencies for developing a hardware and software solution that performs an emergency stop of a UAV when an obstacle is detected based on the domestic single-board computer Repka Pi.

The program started on October 1, with six students of engineering departments of the university participating under the supervision of Vyacheslav Nikitin, PhD, Associate Professor of the Department of Photogrammetry and Remote Sensing. During the training, the team members received the necessary competencies, found the necessary engineering solution and successfully completed the task.

The members of the SSUGT engineering team presented the customer, Mirai Vision LLC, with the developed software for an emergency stop of a multi-rotor drone when an obstacle is detected.

At the presentation of the development, the team members successfully conducted test trials in a simulator with the selected testing algorithm and projection of various situations during a drone flight, including demonstrating an emergency stop of the drone when a wall or a person appears on the way, landing the drone on a flat surface, etc.

– The presence of an obstacle detection system on board an unmanned aerial vehicle is one of the means of ensuring flight safety. Similar systems are widely used in multi-rotor UAVs of the world's leading manufacturers (DJI, Parrot, Autel, etc.). The study and design of means to increase the autonomy and safety of UAVs is an urgent task in the training of engineering personnel, – commented Nikita Andreevich Gurov, a representative of Mirai Vision LLC.

The engineering team members managed to find a suitable solution for the optimal set of sensors necessary for effective detection of obstacles and their optimal location on the drone by analyzing several layout options.

The layout of the sensors ensures the detection of obstacles from 3 directions (under the UAV, above, in front). To solve the problem, different types of sensors were used: rangefinders (laser, ultrasonic, radio frequency), 2D/3D lidars, depth cameras and combined sensor arrays.

Based on a comparison of the parameters of these sensors and taking into account the selected obstacle detection zones (floor, ceiling, wall and person), software modules were developed for ROS (robot operating system) that ensure an emergency stop of the drone in front of an obstacle.

- As a result of the engineering team's work, two problems were solved: software was developed that, thanks to the selection of suitable sensors and their optimal layout, allows the drone to detect an obstacle (person, wall) with an accuracy of up to 95%, and also, in the event of an obstacle being detected through a connection to a Repka Pi single-board computer via the MavLink protocol, stops the drone, - said team mentor Vyacheslav Nikitin.

  • Фото 1
  • Фото 2



Новости
События