Author(s): Jannik Elsaesser; Rocco Palmitessa
Linked Author(s):
Keywords: Biofouling; Environmental monitoring; Predictive maintenance; Survival analysis; LLM
Abstract: Biofouling is the primary cause of maintenance and data gaps for online underwater water quality sensors, which are crucial for environmental compliance. This study proposes a novel method for predicting the likelihood and timing of biofouling events, moving beyond reactive sensor servicing and creating a proactive maintenance strategy. Utilizing a comprehensive survival analysis dataset derived from time series data and service logs from 14 monitoring stations on the Fehmarnbelt Fixed Link project, we applied statistical survival analysis models. Our top-performing model, a gradient-boosted Cox proportional hazards model, demonstrated high predictive accuracy (concordance index of 0.984 and 3-day Brier-score of 0.045). This model is highly promising as a decision support tool to optimize maintenance schedules, potentially reducing costs and environmental footprint while significantly extending sensor uptime.
Year: 2026