Self-Check system of heuristic algorithms as a “New moral” of intelligent systems

• Published in: AIP conference proceedings Vol. 2700; no. 1
• Main Author: Tikhanychev, Oleg V.
• Format: Journal Article Conference Proceeding
• Language: English
• Published: Melville American Institute of Physics 09.03.2023
• Subjects: Algorithms; Autonomy; Errors; Heuristic methods; Intelligent systems; Morality; Robotics; Robots; Safety; Software; Transportation systems
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0124956

The relevance of research subject choice, such as the use of autonomous robotic systems for various purposes and research object, which are algorithmic problems arising in the control of algorithms safety, is determined by the existing contradiction between the need for robotic systems autonomy and the complexity of the software implementation of this requirement. The study was carried out on the example of the application of robotics in the field of autonomous transport systems. The implementation of autonomy trends in this area gives rise to certain legal and technological problems associated with the safety of autonomous systems algorithmic support. A manifestation of such problems are critical application errors, which are difficult to identify at the development stage. In robotics, the main potential reason of these errors is software functioning of autonomous robotic systems outside the framework of the deterministic algorithms that were tested during their development. And if for ordinary technical systems the consequences of application errors are predictable, and the responsibility for their results is determined, then for the case of using autonomous robots, the situation requires a solution. In the article, using general scientific methods of analysis and synthesis, the main aspects of the current state and prospects of the autonomous systems problem use algorithmicization are investigated. On the basis of a general analysis of algorithmic problems of autonomous robots use, the possible consequences and causes of algorithmic errors occurrence that arise during their use, the scientific problem formulation of solving the problem of heuristic algorithms development that determine their application using a dynamically generated behavioral predicate system similar to human morality is synthesized.