Autonomous Planning for Targeted Observation of Severe Weather
摘要
This paper describes the decion-making architecture for an autonomous airborne scientist that closes planning loops over high fidelity weather models. The autonomy architecture exploits: local computing on the aircraft; edge computing in the field; and cloud computing accessible through the Internet, and other sensing and computing at fixed sites. This paper describes strategic-level information-space planning for determining where to fly and when to release air-launched drifters and tactical-level motion planning in complex flow fields.