facebook url
R&D - Artificial Intelligence Technologies
Artificial Intelligence Technologies

In its most general form, artificial intelligence can be defined as the ability of computers to perform certain tasks in a human-like manner. Especially in the last 10 years, thanks to the deep learning methods developed within the scope of machine learning, which is a sub-branch of artificial intelligence, it is seen that computers can perform tasks that are very difficult to fulfill with classical approaches more successfully than humans.

ASELSAN Research Center conducts studies on computer vision, natural language processing and decision making, which are among the most popular application areas of artificial intelligence.

Control with Machine Learning

Intelligent control can be defined as a multidisciplinary set of topics that include algorithms developed to solve control problems by adapting biological intelligence to control systems, replacing a human in control tasks or inspired by the problem-solving capabilities of biological systems.

Machine learning and reinforcement learning can support the solution of problems in the field of control after the amount of digital data, processing power and successful application areas have increased considerably in recent years. In this direction, machine learning-supported, intelligent and adaptive controller algorithms can provide solutions especially for control systems with nonlinear effects, high order modeling, noise or system model variability.

Within the scope of these issues, Aselsan Research Center is working on system recognition, designing controllers according to a standard infrastructure and fine-tuning them as they operate, and optimal control in nonlinear systems.

Swarm Intelligence

The recent rapid developments in computing and communication technologies, and the fact that unmanned systems have started to cooperate with humans and other unmanned systems have led to the emergence of swarm intelligence as a new branch in robotics. This new field has taken its place in the literature as "Multi-Robotic Systems" or as it is called in the artificial intelligence literature "Multi-Agent Systems".

The Multi-Agent Systems robotics sub-branch is more advantageous than single robot systems in terms of simultaneous operation, general system redundancy, cost effectiveness, task flexibility, capacity and diversity.

Within ASELSAN Research Center, pioneering technological studies are carried out with the aim of increasing the competence of ASELSAN and our country in these areas.

Natural Language Processing

Natural language processing is an application area that basically deals with computers understanding words and sentences as humans do, and has sub-branches such as question answering, classification, and command acquisition/transmission. In the field of natural language processing, which has been studied at the Research Center since 2018, basic technology gains have been achieved and various studies are being carried out for ASELSAN products, ranging from voice assistants to automatic content tracking systems, from detecting the relevancies between documents to designing classifiers to improve processes.

Significant progress has been made in the classification and association of text data, and pioneering work is being carried out in the development of a Turkish voice assistant and the prediction of interactions in social communication environments.

Computer Vision

Computer vision is a field of application that is basically about computers having a vision capability like humans and has important sub-branches such as detection, tracking and classification. Studies have been carried out in this field since the establishment of the Research Center. ASELSAN has reached a maturity that will provide ASELSAN products with competitive features. In particular, pioneering studies are being carried out in areas such as augmented reality, operator fatigue detection and visual classification.

Specific solutions are being developed to meet the need for big data, which makes it difficult to implement learning-based approaches, especially with visual domain adaptation techniques. The techniques developed are planned to be used in ASELSAN products in the medium term.

Decision Support and Recommendation Systems

Decision making is basically an application area where computers evaluate situations like humans and make the most appropriate decision. Studies on decision making, which has a wide range of applications from autonomous systems to playing games such as chess and GO, are carried out within the Research Center. Decision-making studies, which seek solutions to resource allocation and highly complex optimization problems, have recently intersected in the field of move determination under high uncertainty.

Considering that the decision support capabilities of deep learning techniques will benefit not only ASELSAN products but also ASELSAN processes, pioneering studies are being carried out on the classification of software errors, which is a part of the software development process. With these pioneering studies, methods that provide results equivalent to human performance without the need for specialized personnel are being developed.

Human-Machine Interaction

With the developing technology, there is a need to develop new methods and supports in order for people to evaluate the increasing amount of digital data and high amount of data in recent years and to make decisions. In this direction, the field of human machine interaction deals with the design of interfaces, control mechanisms and intelligent systems that are necessary to maximize the awareness and control of operators with their natural action and perception mechanisms.

The need to collect this information with minimal delays and to send commands just as quickly has been solved thanks to the acceleration of mobile communication technologies in recent years. On the other hand, the development of virtual reality technology has made a significant contribution.

Studies on these structures, which are considered to make a huge difference in many areas, are carried out within the Research Center, and activities are carried out to acquire new and game-changing technologies.


© 2022 - ASELSAN A.Ş.