September 20-26 2021
Artificial intelligence to help with border controls
During the 1st wave of the COVID-19 pandemic, between April and July 2020, many countries banned non-essential travel so as to limit the spread of SARS-CoV-2. After the 1st wave, many European countries lifted these restrictions. States adopted different border surveillance systems, such as allowing travel only to countries with low rates of the virus, or imposing quarantine on all in-comers. US researchers were able to use artificial intelligence (AI) to help with border surveillance in Greece during the epidemic.
These researchers created an algorithm that enabled dynamic AI functioning. This involved computer servers capable of dealing with complex calculations from enormous data bases, while continuing to learn and to refine constantly thanks to the algorithms (machine learning). All passengers had to fill out a questionnaire prior to their arrival, and the algorithm, named EVA, made a calculation as to the rate of infected persons within the travel group. In addition, after calculating infection risks, EVA was able to identify passengers who should be tested by PCR. To make this calculation, the algorithm did not use personal data, but took into account testing resources in the country of origin.
In addition, EVA was able to draw up a list of high-risk countries. Scientists were able to demonstrate after tests that EVA had been able to detect 1,85 times as many asymptomatic infected people than random controls. The researchers also worked with lawyers, epidemiologists and politicians to ensure the confidentiality of personal information.
In conclusion, EVA enabled better identification of asymptomatic infected people and provided real-time information to help the government. Artificial intelligence may therefore be able to be used in the future to prevent tourism being curtailed during a pandemic, thus avoiding bringing a country’s economy to a standstill.