Cyber Physical Intelligence for Oil Spills (CPI)

Abstract:

The National Academy of Sciences estimate 1.7 to 8.8 million tons of oil are released into global waters every year. The effects of these spills include dead wildlife, oil covered marshlands and contaminated water. Deepwater horizon cost approximately $50 billion and severely challenged response capabilities. In such large spills optimizing a coordinated response is a particular challenge. This challenge can be met in a revolutionary new way by using an objectively optimized Cyber Physical Decision Making System (CPS) for rapid response products and a framework for objectively optimized decision-making in an uncertain environment.

The CPS utilizes machine learning for the processing of the massive real-time streams of Big Data from comprehensive hyperspectral remote sensing acquired by a team of low-cost robotic aerial vehicles, providing a real-time aerial view and stream of hyperspectral imagery from the near UV to the thermal infrared, and a characterization of oil thickness, oil type and oil weathering. The objective decision making paradigm is modeled on the human brain and provides the optimal course trajectory for response vessels to achieve the most expeditious cleanup of oil spills using the available resources.

In addition, oil spill cleanups often involve surface oil burns that can lead to air quality issues. The aerial vehicles comprehensively characterize air quality in real-time, streaming location, temperature, pressure, humidity, the abundance of 6 criterion pollutants (O₃, CO, NO, NO₂, SO₂, and H₂S) and the full size distribution of airborne particulates. This CPS can be readily applied to other systems in agriculture, water conversation, monitoring of stream quality, air quality, diagnosing risk of wild fires, etc..