The integrated optimization of autonomous air and marine platforms is becoming a grand challenge for the efficient utilization, monitoring, and protection of our environment and life on Earth. Applications include research, environmental monitoring, conservation, climate change mitigation, weather prediction, ocean forecasting, transport and distribution of goods, security, air-sea operations, communication, search and rescue, space and marine industry, and the blue economy. To achieve successful integrated air-sea autonomy in such applications, predicting hazards and reducing risks is critical, especially for autonomous vehicles with limited actuation or high costs. Much progress has been achieved in the past decade in either marine or air path planning. Some efforts have included ocean risks or air risks, but integrated air-sea applications are not yet commonplace.

In this work, we apply the MIT-MSEAS general partial differential equations for exact multi-objective reachability and optimal planning to guide autonomous air and sea drones that operate in fastest time in uncertain dynamic environments and steer clear of hazards along their path. For the first time, we combine weather, ocean, and environmental hazard forecasting with dynamic multi-objective optimal control to obtain hazard-time reachable sets, Pareto fronts, and optimal paths. Our first hazard-time optimal path planning application consists of an autonomous air drone that crosses the Atlantic Ocean optimizing travel time while avoiding hazardous rains. The second is the hazard-time optimal transport of an ocean vehicle by an air drone followed by a hazard-time optimal ocean mission. The air drone travels to a target location, drops the ocean vehicle, and the ocean vehicle completes its underwater mission in the fastest time, avoiding hazards. Other collaborative missions are also presented.