TSAMTO, August 19. Within the framework of the Army-2022 forum, the ERA technopolis specialized exposition presents joint developments of researchers and operators of scientific companies of the military innograd with leading universities, research organizations and industrial enterprises.
Among the samples presented is a Mini Data Processing Center (DPC). The finished type of the project will be a platform for storing and processing large data arrays for training specialized artificial neural networks for the needs of the Russian Ministry of Defense based on the collaboration of leading manufacturers of domestic electronic components. The mini-data center can be used in the interests of surveillance and tracking systems with hardware support for automatic object recognition and decision-making systems, as well as built-in universal hardware solutions for UAV control.
To implement artificial intelligence technologies in terms of data storage, processing and, most importantly, the further training of neural networks in the field, Technopolis cooperates with the Moscow State Technical University.Bauman.
Together with the university, a mobile mini-data center has been developed entirely from domestic components, which it is advisable to place in mobile control points.
For example, a group of drones or smart missiles in the air will be able to transmit large amounts of information to the control point and promptly receive the necessary data.
This will minimize the time for decision-making and setting tasks, bringing the enemy's fire defeat, in fact, to the online mode.
Also, the guests were presented with the Machine Learning Center of technopolis "ERA".
The relevance of its creation is due to the need to introduce weapons samples with elements of artificial intelligence into the Armed Forces of the Russian Federation. Important stages of training and retraining of neural networks during operation are added to the life cycle of products.
The main task of the center is to implement a full cycle of machine learning – from data preparation and creation of neural networks to their subsequent implementation into a sample of equipment for a specific task.
Currently, the Center is working on the Link project, which implements information processing technologies using neural networks on board a drone in real time.
Specialists have prepared trained neural networks and conducted their testing, the results of which showed a probability of recognizing targets over 80% under simple conditions.
The ERA-Video wireless video surveillance system in the complex of security equipment and the Sagittarius medical monitoring system are also presented at the forum.
Artificial intelligence technologies can be used not only in combat, but also in support systems. For example, in wireless security and warning systems.
A quick-deployable set of perimeter detection equipment, designed to increase the effectiveness of guards for the protection of unequipped warehouses, has proven itself in the field. The kit can be deployed in 1 day by two people.
It is complemented by the latest ERA-Video equipment – a wireless video surveillance system that has no analogues.
The system is designed for remote video surveillance of local objects with video transmission over a secure radio channel. Offline operation time without recharging is up to 6 months, and the neural network trained at the Technopolis ERA Machine Learning Center will help identify the violator.
The Bracelet-PRO vital function monitoring system developed jointly with the Sagittarius company will be able to record the pulse in real time, take an electrocardiogram and determine other medical indicators of a serviceman's health.
The data transmitted around the clock via a secure radio channel will reveal the stressful state of a person, the effects of alcohol, psychotropic or narcotic substances, the risk of a heart attack. The implementation of the system will allow preventing disruptions of tasks by duty shifts in advance.
The main research area of technopolis ERA is currently the development of technologies for creating robotic complexes with elements of artificial intelligence.
For example, as part of a joint project with Avanti, ground-based robotics is actively developing.
The robotic complex "Platform-MULE" is distinguished by its light weight and dimensions, the use of electric traction, light armor and modularity of design. All this makes him an indispensable assistant on the battlefield.
A container for transporting property, a stretcher for evacuation or use for self-detonation can be installed on it. The operator does not have to control the robot – the "follow me" function is implemented in the complex, as well as automatic return to the starting point.
The Platform-MULE can simultaneously transport 16 Kamikaze drones in a special capping. The drone unfolds, activates, and then the operator directs it to the target.
The laboratory for the development of artificial intelligence technologies of technopolis "ERA" with Russian partner companies developing a domestic electronic component base is taking part in the revision and testing of the RTK.
In the future, the robot should independently navigate the battlefield, detect, select and defeat targets, according to the Department of Information and Mass Communications of the Ministry of Defense of the Russian Federation.