It’s a summer of testing and experimentation across a broad area of Russian farmland. This season major players in the nation’s agricultural industry are field-testing harvest automation based on combines equipped with artificial intelligence and on-board monitors. The results could mean a major new player coming to the table in global precision agriculture.
Unlike most European and North American precision agriculture systems that use GPS-guided vehicle movement and an assortment of laser and photo sensors, Russia’s Cognitive Agro Pilot system guides tractors, harvesters and sprayers with a single high-resolution color video monitor linked to an artificial-intelligence (AI) “Computerized Neural Network” programmed to recognize field and crop borders, crop rows, windrows and nearby obstacles.
The test includes 242 combines operated by Rusagro (the nation’s largest agricultural holding firm) on a variety of farms in the market area of Russia’s state-owned agricultural lender, Sberbank. The auto-drive system is the product of a joint-venture between Sberbank and Cognitive Technologies Group and will see service in four climate zones across the Belogrod, Tambov, Kursk and the Orel regions.
The Cognitive Agro Pilot system is designed to control a combine’s unmanned movement along harvest borders over row-crop and small grain stands, as well as on irregular windrow paths at field speeds up to 15 km/hour (9 mph). Company literature emphasizes the system’s ability to reduce fatigue-related errors for machine operators, many of whom spend 12-hour shifts in the cab.
So far, the combines will still need operators, but Cognitive Technologies officials say the automatic controls will enable them to focus more on managing and controlling other harvest settings of the machines, such as header height and angle, threshing clearances and fan settings for grain cleaning. In the future, however, they say they plan to launch a completely automated harvester operating system.
The company has developed a number of components for self-driving cars, as well as autonomous control systems for rail and tram transportation. Its customers include Russian Railways, Rusagro and South Korean auto parts maker Hyundai Mobis.
Cognitive Agro Pilot is capable of harsh weather operations, and is designed to function in a variety of lighting situations. Because it doesn’t depend on global positioning signals for guidance, it can operate in areas with weak satellite reception. The total system includes an on-board computer, the camera, a display, a set of cables, the control servos, wheel monitors, a dosing meter and other related control equipment.
The technology has undergone testing in the U.S., Brazil, China as well as several Russian agricultural regions, and won Autonomous Vehicle Technology magazine’s international ACES award in 2019.
Rusagro’s chairman Vadim Moshkovich says the system will increase the efficiency of Russian harvests significantly by eliminating costly trip overlaps or missed grain at the edge of the crop being harvested with tolerances held as close as 20 centimeters (8 inches).
In general, Rusagro expects significant reductions in grain losses and looks forward to 3-5% reductions in overall costs in grain production with the use of the auto-guidance system.
“Our strategy is aimed at increasing per-hectare economic returns, and digitization is the only way to achieve that goal as we scale such improvements throughout agriculture,” says Roman Shkoller, CEO of Rusagro’s Agricultural Business Division. “Interestingly, the large-scale adoption of the system to so many harvesters comes in a year when Russia is expecting record wheat crops.”
Cognitive Technologies recently announced its launch of the first open-source AI databank stocked with thousands of clips of video scenarios generated globally in its test program for software developers working on agricultural guidance systems. The company also told Reuters in February it is considering stock offerings sometime in the next 3 years.
Chief Executive Olga Uskova says an initial public offering (IPO) is likely sometime after 2023 as Cognitive Technologies seeks to be part of global efforts to automate transportation. The company is eager to increase its capacity and mass-produce its technology.
“The first IPOs in this sector will be highly successful, and it’s important not to sleep through the moment,” Uskova explains.
Industry watchers say while there is obvious potential for industrial growth around transportation and industrial vehicle automation, independent firms such as Cognitive Technologies will face stiff competition from large automotive firms such as General Motors, Toyota and Tesla — which are working on proprietary in-house autonomous-driving systems.