Tracking multiple targets at sea: not a question of if.

With the ever increasing world reliance on large cargo containers, it would seem that tracking multiple targets like 20/40 feet containers should be an on-going concern. The most recent occurence is that of the MSC Napoli close to the English coast.

While the modeling of Marc le Boulluec at Ifremer on the hydrodynamics of drifting containers is a good start [the study reminded me of the issue of space debris falling back into the atmosphere and the probable need to go probabilistic on these laws],

it is also important to realize that the tracking of multiple targets is a difficult issue that includes data fusion especially if one uses different data acquisition systems (satellite imagery, airplane sightings, human eyewitnesses.....) at different times. The other issue is finding out when these drifting containers are not a danger anymore as they are not designed to stay afloat but will certainly wreck any unsuspecting small boats
In line with that, I would be interested to see how one has summarized the knowledge on head-sea parametric rolling and integrated it into an algorithm used for predicting trajectories. I hope it goes beyond a simple formula.