Phenomena of extreme behaviour of large engineering objects capture the attention of the public-at-large as they can incur loss of life, intolerable damage to the environment and severe functional disruptions of vital activities for our organized societies. The consequences accruing from a rare accident can, in some cases, be so enormous that they outplay the low probability of occurence of the triggering event.

Our group is focused on the understanding of extreme engineering phenomena that arise in the marine/maritime world; and on the creation of efficient methods for their prediction and assessment. We use mathematical modelling, advanced nonlinear systems investigation techniques and, when possible, physical model testing. Our practical goal is to contribute to the establishment, from first principles, of sound safety assessment methods suitable for ships and other floating objects, which can be effectively applied in design and regulation development.

Excitation by unforeseen high loads and faults or omissions in a system’s design or operation, are the commonly identified factors behind exceedences of safety limits. However, and unlike general perception, the realization of an unsafe state could also be the outcome of a complex, poorly understood, nonlinear process. Effective suppression of dangerous events by prescriptive or even performance-based regulations can hardly ever be fully warranted; because new hazardous occurrences do not always relate to experience. Implementation of economies of scale sometimes compels design innovations beyond the realm where understanding of a system’s behaviour is satisfactory. As a result, from time to time, gaps of knowledge come to surface, recognized sadly after a devastating accident. This seems to be an across the board reality of our modern technological world.

The investigation of rare extreme phenomena can motivate significant cross-disciplinary scientific advances. Sometimes, the underlying basic dynamics is common to several systems appearing, at first sight, unrelated.