Research on cybersecurity methods for robotic complexes under connection to industrial Internet of Things networks

Abstract

The expansion of the industrial Internet of Things has turned a robotic complex from an autonomous controlled object into a node of a distributed network, where actuators, controllers and sensors exchange data through public protocols and edge gateways. A contradiction arises: typical perimeter defences are designed for office information systems and poorly account for the strict real-time constraints inherent in the control loop. The aim of the work is to systematise the protection methods for such a complex and to quantitatively assess their combined effectiveness when connected to an industrial network. The study applies decomposition of the attack surface across architectural layers, mapping of threat classes to countermeasures, and an author's model of an integral cyber-resilience indicator with sensitivity analysis of key parameters. The calculation shows that the largest gain in security comes not from reinforcing a single barrier but from a coordinated combination of network segmentation under the zones-and-conduits model with behavioural detection of anomalies in control actions; when the share of protected end devices reaches 0.9, the integral coefficient attains 0.79 against 0.46 in the baseline configuration. The sensitivity of the indicator to segmentation depth declines after the third isolation level, which sets a practical limit on network partitioning. The obtained dependence means that the cyber resilience of a robotic complex is determined not by the range of installed tools but by the completeness of their coverage of the control loop and by the consistency of incident response.