Hybrid technology with AI components for construction in extreme conditions: models of adaptation and decision-making in offline mode

Abstract

The research focuses on the development and implementation of hybrid technological solutions with integrated artificial intelligence components for construction in extreme environments. Extreme conditions characterized by high uncertainty, dynamism and unpredictability require innovative approaches to the organization of construction processes. The paper proposes a conceptual model of a hybrid technical system with AI components capable of operating offline, adapting to changes in the external environment and making decisions in conditions of limited information and resources. Based on a comprehensive analysis of modern research, a multilevel decision-making system architecture has been developed that integrates symbolic and machine methods of artificial intelligence. A series of experimental studies has been conducted in simulated extreme conditions, including Arctic regions, highlands, and areas with increased seismic activity. Empirical results demonstrate a significant increase in the efficiency of construction operations: a 37.8% reduction in critical task completion time, a 29.5% reduction in resource costs, and a 42.3% increase in adaptability to unforeseen changes in conditions compared to traditional approaches. The proposed methodology opens up new prospects for the development of intelligent construction technologies capable of operating in extreme conditions with minimal human intervention, which is of particular importance for strategically important and hard-to-reach infrastructure facilities.