Un capteur connecté envoie des impulsions laser de haut en bas sur la longueur de chaque câble, analysant la façon dont le signal lumineux a changé entre l’envoi et le retour. Si les larves se nourrissent dans l’un des stipes, les vibrations de leurs bruits de grignotage traverseront le parenchyme et modifieront la fréquence des signaux que le capteur détectera.
Prof Boon S. Ooi formé à Glasgow conduit une équipe de scientifiques de l’Université des sciences et technologies du Roi Abdallah d’Arabie saoudite (KAUST),
Les perspectives sont très intéressantes particulièrement en termes de coût elles devront être confirmées.
Important et rare, un site technologique francophone a relaté cette publication.
Responsible for millions of dollars in crop losses annually, the red palm weevil is a flying beetle that lays its eggs inside date palm trees. It could soon be possible to detect such infestations earlier than ever, using fiber optic cables.
While the red palm weevil beetle itself feeds on date palms, the problem really begins once its eggs hatch, and all the larvae also begin munching on the tree. Typically, by the time that affected date palms show visible signs of distress, it’s too late for them to be saved.
Several methods have been suggested for the earlier detection of infestations – these have included the use of sniffer dogs or the insertion of sound probes into the tree’s trunk, the latter of which detect the chomping sounds of feeding weevil larvae. According to scientists at Saudi Arabia’s King Abdullah University of Science and Technology (KAUST), however, such techniques aren’t practical for use on large-scale date farms.
Led by Prof. Boon S. Ooi, a team from the university has instead looked to long fiber optic cables that can be wrapped around the trunks of multiple trees.
An attached sensor sends laser pulses up and down the length of each cable, analyzing how the light signal has changed between being sent out and coming back. If larvae are feeding within any of the trees, the vibrations from their munching noises will travel through the wood and alter the frequency of the signals, which the sensor will detect.
In the development of their experimental system, the researchers started by recording the feeding sounds of 12-day-old weevil larvae, along with typical background noises such as birds and wind. They then developed an algorithm that identifies and amplifies the frequency-changes associated with the feeding sounds, while discarding the others. Additionally, if weevil-associated changes are detected, signal-processing techniques are able to determine the point along the cable at which they’re originating.
When lab-tested on small trees, the technology was found to accurately differentiate between healthy trees and those that were infested. The scientists now estimate that once developed further, the system could cheaply and continuously monitor about 1,000 date palms at once via a single 10-km (6-mile)-long cable.
A paper on the research was recently published in the journal Scientific Reports.
Source : KAUST via AlphaGalileo