Home Robotics Unmanned tractor tested in Russian fields

Unmanned tractor tested in Russian fields

by admin

Unmanned tractor tested in Russian fields
Russian company Cognitive Technologies conducted in Tatarstan the first tests of unmanned tractors with computer vision system of its own design. The hardware-software complex is planned to be installed not only on tractors but also on other agricultural machines such as harvesters, seeders and so on, write "Izvestia.
Thanks to unmanned transport and computer vision systems, Russian farms will be able to significantly increase their profitability, because now expensive machinery is used inefficiently. For example, a combine harvester works for an average of six hours a day, and the rest of the time it is idle. Unmanned tractors and harvesters will work in the fields day and night, 24 hours a day. They won’t be hindered by poor visibility, rain or fog, and even in pitch darkness, an unmanned tractor can mow the grass or pick potatoes as effectively as it does during the day.
Unmanned agricultural machinery in Russia will be promoted by a new agricultural holding created by Cognitive Technologies with the manufacturer of agricultural machinery Rostselmash and the Soyuz-Agro agricultural holding on the basis of the Innopolis special economic zone.
Night vision cameras can be used for unmanned systems in total darkness, but more effective is the use of lidar in conjunction with a pre-marked map of the area that is stored in the tractor’s navigation computer. U.S. colleagues recently tested an unmanned vehicle system vehicle in complete darkness – and determined that lidar with a navigator drives a car much safer in the dark than a human driver.
During a test of an unmanned Ford at a test site in Arizona, a 3D map of the terrain was produced, and in the dark, lidar used laser pulses (2.8 million pulses per second) to plot the path, as if probing the road with an orientation to the 3D map.
Unmanned tractor tested in Russian fields
There are no plans to equip the Russian tractor with lidar yet, because this would significantly increase its cost. At the moment, according to the developers’ estimates, the cost of the hardware-software system for unmanned vision is no more than 15% of the tractor’s cost. For Russian collective farms and farms, this is quite an acceptable markup. If a tractor is equipped with a full-featured lidar, the cost of the equipment will increase considerably.
Russian tractors will be equipped with computer vision hardware and software, which includes a stereo pair – a system of two video cameras that shoot video with Full HD resolution, i.e. 1920×1080 pixels.
In addition to video cameras, the Russian unmanned tractor is equipped with GLONASS and GPS navigation and inertial sensors and a computing unit (computer).
"The computer vision system makes it possible to detect dangerous objects with high accuracy, determine their dimensions and coordinates for high-precision mapping, says Olga Uskova, president of Cognitive Technologies. – Thanks to the exact knowledge about the position of the objects in the field, it is possible to remove many of them before the harvest, when they could be a real threat to the mechanical parts of the farm machinery.
A digital map of the field and the mapping of surrounding objects (poles, stones and other artifacts) is done during pre-seeding operations such as fertilizing and harrowing. The operation takes place in the spring, when the crops are not sown, so that all artifacts are captured well by the computer vision system. Using these digital maps, the tractor will navigate in the summer and fall, during harvest. It will try to avoid objects that could not be removed from the field in the spring.
The developers are confident that the vehicle’s sensors and computer vision system will be able to detect even such obstacles that are not on the map. It is stated that the system recognizes objects ranging in size from 10-15 cm at a distance of up to 15-20 m. Anything else is safe as long as the stone catcher is installed on the tractor.
Tractor tests with Cognitive Technologies computer vision system
Tests of the unmanned tractor will last about a year and a half. According to Olga Uskova, this is enough time to bring the computer vision system to the stage of industrial use and put it into commercial operation. The software and hardware complexes will be sold to farmers in Russia, as well as in the near and far abroad countries.
Experimental developments in this area are not only in Russia. For example, the Japanese Minister of Agriculture recently raised the topic transition to robotic technology in agriculture due to an aging population, including farmers.
The average age of Japanese farmers is 67, and two-thirds of them are over 65. Russia does not yet have an aging problem because of low life expectancy, but in time it may appear.
Japanese Kubota Corporation is ready to produce unmanned tractors.
Unmanned tractor tested in Russian fields
Prototype of Kubota autonomous tractor with lidar
Small unmanned tractors are also made in Belarus Surely the Russian computer vision system can also be installed on the popular Belarusian tractors produced by the Minsk Tractor Plant.
In the imagination of Russian developers, agriculture in Russia in the future looks like this : "The work of all smart agricultural machinery will be controlled by a minimum number of people from a control center, from where all machines will be monitored. If necessary, he will be able to set the program and time of the tractors, and will not have to be constantly at the console, " says Olga Usova.
In general, approximately as predicted by the writer Nikolai Nosov in his science fiction work "Dunno in Sun City".
Unmanned tractor tested in Russian fields
Independent experts believe that a fully automated future in agriculture is not yet possible. A person must be behind the wheel of the vehicle to monitor the operation of automation and be ready to take control in case of an emergency.
By the way, about its developments in the field of machine vision, Cognitive Technologies in more detail described in a corporate blog on Habrahabr As part of a joint project at the Department of Cognitive Technologies at MIPT in 2014, a robot car was developed that "processes a video stream in real time, recognizes the surrounding scene, detects objects, and generates a control action to solve the task at hand." The robot was tasked with chasing a red ball and pushing it with its bumper.
The video was taken during the report at the fall conference of MIPT young scientists, right in the corridor of the main building.

You may also like