Researchers have developed a portable surveillance device that uses machine learning to detect coughing and crowd size in real-time. The system can use the data to directly monitor flu-like illnesses, community trends, and virus spread potential. The system could help monitor the spread of infection and provide a public health response during an epidemic.