MOSCOW, February 5. /TASS/. Russia’s Geoscan company will launch mass production of an upgraded version of the Pioneer Mini 2 educational quadcopter this spring, featuring support for neural networks and computer vision. A prototype of the drone was presented at the 2nd International Science and Technology Forum ‘Robotics, Intelligence of Machines and Mechanisms’ (RIMM-2026).
"[The upgraded quadcopters] should enter [mass production] this spring," a company representative told TASS.
According to the quadcopter presentation, the drone is equipped with an 8-core processor with a neural accelerator for processing images, swarm flights, and artificial intelligence. Using Python libraries developed by Geoscan, users can integrate their own neural network elements to create their own drone control scenarios.
The drone has a maximum flight time of 11 minutes with a maximum horizontal flight speed of 15 km/h. The quadcopter weighs less than 150 grams.
The Geoscan Pioneer Mini 2 was created specifically for STEAM education. The term STEAM (Science, Technology, Engineering, Arts, and Mathematics) is used to collectively describe the academic disciplines most strategically important for economic development. The drone allows students to explore computer vision algorithms, neural networks, image processing, and autonomous flight programming. Students will also be able to develop and integrate their own payloads for handling of applied problems.
The Pioneer Mini 2 is an upgraded version of the entry-level quadcopter from Geoscan’s educational UAV line. In the summer of 2025, Geoscan’s press service reported that the company would begin delivering Pioneer mini 2 quadcopters with onboard artificial intelligence to educational institutions in the summer of 2025.