Modern approaches to understanding the autonomy of technical systems

Abstract

Now, in various fields of human activity, so-called autonomous technical systems are becoming relevant, capable of providing management with minimal human participation or without human participation at all.

At the same time, some publications interpret the concept of an autonomous technical system capable of making decisions independently in different ways. There are "jargon", from the point of view of the engineering community, the concepts of "Autonomous", "Self-Driving", "Driverless", "Unmanned", "Robotic". In particular, the use of the term "autonomous" leads to ambiguity in the understanding of the concept of "autonomous vehicle", since it remains unclear whether such a vehicle needs any communication or cooperation with external entities for its functioning.

Even greater difficulties arise when translating Western texts or when initially writing Russian-language texts on this topic, since various English-language terms, including those listed above, appear in Russian articles in the form of a single concept of "autonomous".

This article tries to identify the answers to the following questions:

- How is the concept of "autonomy" revealed in generally recognized international standards?

- What basis is laid in the taxonomy of autonomy and how is autonomy measured?

- What taxonomy options are available and what is their rationale?

- What is the connection between autonomy and machine intelligence?

- What is the relationship between autonomy and the capabilities provided by the machine?