Impact of robotic maintenance on environmental safety of production sites and reduction of accidents in specially protected natural areas and Arctic zones

Abstract

The study examines the impact of robotic maintenance on environmental safety and accident reduction during the operation of oil and gas infrastructure in the Arctic and in specially protected natural areas, where extreme conditions and the human factor account for the majority of risks. The goal is to quantitatively assess the effect of robotization and demonstrate its role as a necessary condition for sustainable operation. Methodologically, the research employs a systemic and comparative approach based on statistics from 2010–2023 (over 350 incidents), life-cycle simulation modeling (10 years), and analysis of two models: traditional (manual maintenance and repair) and robotic (UAVs with LiDAR and thermal cameras, ROVs/AUVs, mobile ground platforms, robotic manipulators, remotely operated shut-off valves). Technical and environmental metrics were evaluated: incident frequency per 10,000 hours, severity coefficient, proportion of human errors, average spill volume per incident, contamination area, response time, as well as operational and economic indicators (MTBF, maintenance costs, downtime losses) using STATISTICA and MATLAB. The results showed a multiple increase in safety: for main pipelines, incident frequency decreased from 0.087 to 0.019 (by 4.58 times); for subsea complexes — from 0.345 to 0.052 (by 6.63 times). The share of incidents due to human error dropped from 59.6–81.2% to 5.4–14.5%, while the severity coefficient decreased by 4–5 times. Environmental effects include a reduction in the average spill volume: for pipelines — from 12.75 to 1.15 m³ (by 11.09 times), for pumping stations — from 8.50 to 0.88 m³ (by 9.66 times); response time decreased from 8.5 to 0.7 hours (pipelines) and from 12.8 to 1.1 hours (offshore platforms), along with a proportional reduction in contamination area. Operational efficiency is reflected in an increase in MTBF from 0.962 to 0.994, a decrease in maintenance costs from 155.8 to 88.5 thousand units/year per asset, and a reduction in downtime losses from 18.75 to 2.15 million units/year. A strong correlation (r>0.85) between robotization level and integrated safety index was found, with each 1% increase in robotization reducing risk by 1.2–1.5%. The conclusions confirm the multiplicative effect of robotization achieved through eliminating humans from hazardous zones, predictive monitoring, and early intervention. Recommendations include updating regulatory requirements and developing AI algorithms, decision support systems, cybersecurity, and reliability of robotic systems.