Courier drone is taught to plot a route without GPS

Courier drone is taught to plot a route without GPS

TSU and the Geoscan Group of Companies have signed an agreement on collaboration to create intelligent control systems for drones. In particular, as part of the Smart Campus project, the partners will solve technical problems to ensure the autonomy of a drone even without a GPS signal, the accuracy of cargo delivery, and minimal errors during landing.

- TSU’s partner, Geoscan, has provided the university with a new platform that has high functional capabilities, - says Stanislav Shidlovsky, dean of the TSU Faculty of Innovative Technologies. - In particular, this drone can operate at low temperatures, down to -40°С, and its carrying capacity with absolute flight stability, including in windy weather (up to 15 m/s), is sufficient for a wide range of tasks. The advantage of this flight platform is its adaptation to performing tasks in the climatic conditions of our country, which cannot be said about the DJI platforms actively used in the world.

Now scientists at the TSU Faculty of Innovative Technologies are developing an intelligent superstructure for the drone: they are solving technical problems. Among them are problems that arise in the last mile.

The computer brain of the flying courier transfers all of its work to automatic mode. The difficulty lies in the delivery from the sorting center to the consumer's doorstep. Many large global firms and retailers employ drones, but a person still goes the last mile to the customer. One of the serious technical limitations is the presence of an unstable GPS signal or its temporary absence.

TSU Faculty of Innovative Technologies staff are creating complex algorithmic solutions and orientation systems that will help the drone to independently plot a route using a map of the town. Vision systems, which are also being developed at the faculty, will help the courier drone avoid tall buildings, wires, and other obstacles.

Along with this, landing accuracy is being worked out. Most developers have an error of about five meters. To reduce it, you can use the expensive DRTK GNSS system, which, using an additional ground station, sends corrections to the GPS coordinates to the drone.

- We decided to use a fundamentally different approach, - explains Stanislav Shidlovsky. - The landing sites were equipped with infrared beacons, which are visible for the drone in any weather and serve as an excellent reference point. In the future, landing zones with infrared beacons can become part of any urban area.

At the end of March 2020, the delivery drone passed its first successful tests on the TSU campus: it delivered several cargoes, including a monograph on robotics, from the main building of the university to the student residential complex Parus.

In August, the Faculty of Innovative Technologies employees conducted a new series of tests (VIDEO), during which an air courier carried cargo from the main building of TSU to the Parus and Mayak dorms, as well as to the TSU Institute of Military Education.