When a natural disaster strikes, government agencies quickly activate a decentralized and often disparate network of response teams to collect, analyze, and operationalize all types of data and information. However, current data management strategies and traditional epidemiological methods are not well-equipped to deal with the growing complexities of today’s disaster landscape.
Johns Hopkins University’s Applied Physics Laboratory (APL) http://www.jhuapl.edu is now looking to tackle the challenges of disaster response according to APL’s Jeffrey D. Freeman, Assistant Program Manager for Health Surveillance.
Freeman sees disaster health at APL centering on developing an intelligent and autonomous system to enable context-dependent situational awareness for emergency response and recovery during disasters and other public health emergencies.
APL’s goal is to include full automation of data intake, rapid analysis and visualization of heterogeneous data which includes free text, structured text, images, video, audio, radar, GPS, plus other data sources.
“Advanced machine learning methods such as deep learning, interoperable and autonomous data systems, and non-traditional data sources such as drone and LIDAR sensing can contribute enormously to solutions. Core data functions could also be supported by an APL developed artificial intelligence suite of more than 200 machine learning algorithms”, reports Freeman.
“Recent challenges following the Ebola outbreak in 2014 and the hurricanes in 2017 demonstrated that there are many opportunities where we can help to improve both preparedness for or respond to events to restore the health of a community in a timely manner,” said Martina Siwek, APL’s Health Surveillance Program Manager.