54 AI MAGAZINE
At the execution level, agents suggest different
supporting-supported options across a number of
interdependent tasks in competing task graphs
(representing different missions), taking into account
uncertainty in weather forecasts, intelligence, and
asset and network status. At platform and warfare
area levels, agents optimize subplatform-to-task
allocations. Other agents compute mission context–
dependent values of information and decisions
and manage the flow of information among deci-
sion makers. Details may be found in Han et al.
(2014) and references therein.
Conclusion and Future Work
In this article, we demonstrated the use of Io T and
AI for maritime decision making by identifying
underlying contexts where different planning
(a) Original cases showing temporal evolution (indicated by color gradients) of the probability of smuggler activity in the eastern
Pacific Ocean and the Caribbean Sea. (b) Target corridors (white rectangles) and uncertainty associated in red. (c) Ideal weather
scenario with sweep width = 20 nm. (d) Degraded weather conditions affect the performance of sensors and sweep width = 2 nm
(Mishra et al. 2017b).