A human-centered artificial intelligence framework for satellite-derived bathymetry and S-100 digital nautical chart production

Usluer H. B.

6th GMC 2026-Global Maritime Congress 2026, İzmir, Türkiye, 28 - 29 Eylül 2026, ss.1-12, (Tam Metin Bildiri)

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Basıldığı Şehir: İzmir
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.1-12
  • Galatasaray Üniversitesi Adresli: Evet

Özet

In the transition from hydrographic products designed for safe navigation to the S-100 production and usage ecosystem, the production and international integration of the marine component of the geospatial dataset in accordance with standards is of paramount importance. Aligning observations of water areas around the world with advancements in artificial intelligence allows satellite-derived bathymetry (SDB) to provide scalable, high-quality data not only for deep waters but also for shallow waters.Integrating AI-generated hydrographic data into nautical publications, which are critical for navigational safety, presents challenges such as uncertainty regarding compliance with standards, explainability compared to classical methods, data quality compared to archival data, and accountability. This study describes a human-centered artificial intelligence (HCAI) framework to support the integration of AI-assisted hydrographic data into S-100-compliant digital nautical charts.The system combines a wide spectrum of satellite imagery, machine learning, and deep learning with marine science data to perform uncertainty analysis, while verifying data transformation compliance with S-100 standards through the assistance of explainable artificial intelligence and the participation of a hydrograph expert. Within this framework, the fundamental principle is to attempt to exemplify a structured human-machine partnership using artificial intelligence under human control. The system automatically estimates depth data from hydrographic data and includes checkpoints for reliability assessment, anomaly detection, expert review, and approval, separate from acceptance of the hydrographic dataset's use.The depths obtained from the SDB can be integrated into IHO Standard S-102 Bathymetric Surface products and, following expert verification and feature compilation, into S-101 Electronic Navigation Chart workflows. The study, which also summarizes operational requirements for resource, uncertainty communication, human control, and auditability across hydrographic surveys, supports the integration of S-100 standards using artificial intelligence applications in surface observation while maintaining a professional hydrographic assessment system. Contrary to popular belief, artificial intelligence is not replacing the hydrographer in the production of maritime geospatial data sets, which are critical for navigational safety, but rather promoting digital transformation by working alongside experts.