Automatic traffic scenarios generation for autonomous ships collision avoidance system testing

The Collision Avoidance (CA) system constitutes a key enabling technology for the Maritime Autonomous Surface Ships (MASS), appropriate functionality of which is critical assuring navigation safety. Although several techniques including testing collision scenarios in virtual environment can be employed, trust phase results depends on number tested and their coverage. This study aims at proposing systematic automatic process generation traffic that employed CA testing. First, range investigated parameters defined, samples potential are generated using Sobol sequences. Subsequently, hazardous identified from initially by predefined rules. For these scenarios, risk vector considering weather conditions calculated. A clustering algorithm to identify groups encountered based each scenario COLREGs scenarios. groups, riskiest provided as input test cases development, thus, simplifying selection applied theoretical Short Sea Shipping autonomous vessel, whereas derived discuss advantages disadvantages developed process.