Jonathan Eichild Schmidt

Jonathan Eichild Schmidt

Auto-generated profile, based on publicly available information

Jonathan Eichild Schmidt

Department of Electrical and Photonics Engineering, Technical University of Denmark, Kongens Lyngby, Denmark

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
12 May 2025
Verified
Book Chapter
Closed access
PDF file
GHGaladrielle Humblot-Renaux
Galadrielle Humblot-Renaux
AJAnders Skaarup Johansen
Anders Skaarup Johansen
JSJonathan Eichild Schmidt
Jonathan Eichild Schmidt
AIAmanda Frederikke Irlind
Amanda Frederikke Irlind
NMNiels Madsen
Niels Madsen
MPMalte Pedersen
Malte Pedersen

1 October 2025 in Journal of Physics: Conference Series
Verified
Conference Paper
Open access
JSJonathan Eichild Schmidt
Jonathan Eichild Schmidt
BRBørge Rokseth
Børge Rokseth
Roberto GaleazziRoberto Galeazzi
Roberto Galeazzi
Vision-based perception systems, operating in the visible and infrared spectrums, are increasingly integrated into autonomous marine navigation to enhance situational awareness. However, their reliability in detecting and classifying targets is highly sensitive to environmental and operational conditions, raising critical concerns about trustworthiness – especially when system outputs are used to support human or AI-based navigation decisions. Ensuring operational safety demands robust methods to evaluate when such systems can be trusted in real time. This paper proposes a novel framework for online assessment of vision-based perception systems in maritime environments. The method enables real-time performance monitoring under dynamic and challenging conditions by analyzing sensor data in the context of changing environmental factors. A key feature of the framework is a degradation metric that signals reduced system reliability, which can be used to alert onboard or remote operators when perception outputs should no longer be relied upon. We validate our framework using real-world data collected in Danish coastal waters, demonstrating its ability to detect data quality degradation and identify conditions where vision-based systems may provide unreliable outputs. By enabling selective trust and timely intervention, the proposed approach supports safer integration of vision-based technologies in autonomous and assisted ship navigation, while mitigating information overload for human operators.