The use of artificial intelligence and automatic remote monitoring for mosquito surveillance

Abstract

Mosquito surveillance consists in the routine monitoring of mosquito populations: to determine the presence/absence of certain mosquito species; to identify changes in the abundance and/or composition of mosquito populations; to detect the presence of invasive species; to screen for mosquito-borne pathogens; and, finally, to evaluate the effectiveness of control measures. This kind of surveillance is typically performed by means of traps, which are regularly collected and manually inspected by expert entomologists for the taxonomical identification of the samples. The main problems with traditional surveillance systems are the cost in terms of time and human resources and the lag that is created between the time the trap is placed and collected. This lag can be crucial for the accurate time monitoring of mosquito population dynamics in the field, which is determinant for the precise design and implementation of risk assessment programs. New perspectives in this field include the use of smart traps and remote monitoring systems, which generate data completely interoperable and thus available for the automatic running of prediction models; the performance of risk assessments; the issuing of warnings; and the undertaking of historical analyses of infested areas. In this way, entomological surveillance could be done automatically with unprecedented accuracy and responsiveness, overcoming the problem of manual inspection labour costs. As a result, disease vector species could be detected earlier and with greater precision, enabling an improved control of outbreaks and a greater protection from diseases, thereby saving lives and millions of Euros in health costs.