Session: 06-17-01 AI Technology for Ocean Engineering

Submission Number: 156343

Towards a Systematic Safety Evaluation Framework for Mass in Congested Waters 

A clear trend towards the maritime autonomous surface ship (MASS) has been observed within the maritime industry in recent years. MASS is designed to ensure safe, reliable, and efficient operations across various environments. One key ability to achieve full autonomy is making decisions and taking actions independently by the ship system. However, as the complexity of system design and functionality evolves, the hazardous events related to the operations could be more complex and emergent, especially in congested waters that exhibit more complicated traffic situations. Risk awareness and safety evaluation are proactive manners to advance safe and intelligent autonomy for the MASS system. Therefore, this study presents a systematic safety evaluation methodology for MASS in congested waters, aiming to enhance its safety performance while adhering to the International Regulations for Preventing Collisions at Sea (COLREGs). The developed framework aims to address risk perception and risk assessment for MASS. Specifically, a systems theoretic process analysis (STPA) method is applied for risk identification from the control perspective for the autonomous navigation system for MASS, with the objective of evaluating hazardous events, unsafe control actions, and potential causes. The outcomes of STPA are further incorporated into a Bayesian network (BN) for estimating the current risk level at specific spatiotemporal points within the mission area. The proposed methodology is validated through a simulated case study within the Singapore port area. Results demonstrated that the integration of risk analysis methods with the autonomous control system supports the identification of potential hazards associated with operations and their causal factors, all of which could be further utilized to guide decision-making and adaptive behavior to enhance the autonomy degree and system safety. The proposed framework in this study is tailored for the MASS system, but it also can be adapted to other marine autonomous systems.

Presenting Author: Xi Chen Technology Centre for Offshore and Marine, Singapore (TCOMS)

Presenting Author Biography: Dr. Xi Chen is currently a research scientist from Technology Centre for Offshore and Marine, Singapore (TCOMS). Her main research interests cover risk analysis and safety-based intelligent decision making for marine autonomous systems. She obtained the Doctoral degree from Memorial University of Newfoundland, Canada, and obtained Master and Bachelor degrees from China University of Petroleum (East China).