A novel usage of rough sets in design of data fusion systems





Digital Twin, Machine Learning, Design Uncertainty, Wave Forecast, Safety Prediction


Design changes for crewless vessels are unexplored compared to the maritime design processes that have been utilized and updated for hundreds of years. This paper presents an exploration into how autonomous and unmanned systems can impact maritime design, specifically focusing on how well they can fuse multiple types of information. Currently, formal and informal communication onboard crewed vessels between various departments is critical in constructing a view of the vessel’s current health and future capability. A
major focus area is determining whether utilizing data classification techniques can replace these human centered decision processes, and what the design implications of losing the human synthesis will be. This paper proposes a mechanical spring-mass-damper system with base excitation using real-world ocean data to be used to perform analyses. Rough Set Theory (RST) is a data classification technique that can be used for the characterization of a set of objects, finding dependency between attributes, and creating rules for
making decisions. RST is compared with other data classification techniques to determine where each classifier succeeds and how they can generate information useful in design. By integrating the results of these analyses, this paper identifies ways to begin fusing multiple information types and how this will impact marine design in the future of crewless systems.




How to Cite

Sulkowski, B., & Collette, M. (2024). A novel usage of rough sets in design of data fusion systems. International Marine Design Conference. https://doi.org/10.59490/imdc.2024.825

Conference Proceedings Volume


Conference papers