New Surveillance Tools – Not A Panacea, But Technologies Are Getting Refined | Homeland

Technological advances in plant protection are such that before long, almost no spore or insect will escape the inquisitive eye of multispectral cameras and other optical readers that will examine the fields. Indeed, the tools are being refined thanks to various research projects and tests carried out on agricultural businesses. Although they cannot replace the agronomist, they promise to offer good helping hands to agricultural producers.

Automated Traps

The device looks like a letterbox like you can find in the thousands along the country roads of Quebec. Add to it a hole in the shape of a chimney, one or more miniature cameras concealed inside the box, with in addition, installed on its interior bottom, a film intended to stick the insects there, and you obtain one of the models of automated trap for insect pests currently installed in the fields of the province.

“For several years, we have been working a lot with automated traps,” says Julien Saguez, entomologist and biomonitoring researcher at the Grain Research Center, CÉROM, whose efforts focus on three species of butterfly pests: cutworm black armyworm, armyworm and western bean cutworm. “Usually, an insect monitoring network is simple: they are traps that contain a pheromone that you put inside and that will attract insects,” says the CÉROM entomologist, whose network monitoring has 30 automated traps scattered throughout Quebec.

What differentiates the automated trap from its classic version is a camera that can detect the presence of insect pests in a field, identify them and count them, without having to move. The information is then sent via Wi-Fi, directly to a phone or computer. “It can allow us, sometimes, to anticipate insect detection in the field to know where and when they are present”, underlines Julien Saguez, who recalls the importance of intervening early in the spreading of insecticide in order to limit the damage that critters can do.

Photo: Archives/TCN

If the technology promises, it still requires improvements, however warns Jean-Philippe Légaré, entomologist at the laboratory of expertise and diagnosis in phytoprotection of the MAPAQ. The ability of systems to recognize insects remains an important issue. “For large insects, the recognition algorithms developed by companies work quite well,” observes the entomologist, whose experiments on smaller insects still show significant limitations. “In the cruciferous, for example, we tested the swede midge, which is a very, very, very small insect, and at that time there are many more misidentifications that occur” , says the specialist.

The systems currently being developed, partly thanks to artificial intelligence, however, have the advantage of learning from their mistakes. “Some types of traps use deep learning technology [deep learning] which improves the information transmitted to the system so that it can refine its ability to recognize insects,” explains Jean-Philippe Légaré. “You tell the program where it was right and where it was wrong, and with that information, the program improves its learning and becomes more and more reliable. As many intruders can enter the trap, it is important to properly recognize the insect we are dealing with in order to make the right decisions”, specifies the MAPAQ spokesperson.

Butterfly counting operation inside the Trapview automated trap. The butterflies are targeted and the information processed by the algorithms makes it possible to identify them.

Butterfly counting operation inside the Trapview automated trap. The butterflies are targeted and the information processed by the algorithms makes it possible to identify them.

Optical readers for the fight against beetles

To this kind of facial recognition (dorsal, should we perhaps write, since the backs of insects often allow them to be better identified) applied to insects, is added at least one other technology developed by the technological research center of Quebec INO, in collaboration with potato producer Patates Dolbec, from Saint-Ubalde, halfway between Quebec City and Trois-Rivières. “We have a project that is starting to detect the presence of beetles on potato plants,” explains François Châteauneuf, director of the Sustainable Resources, Agriculture and Infrastructure business unit at INO. Thanks to optical readers, installed directly on the machinery and linked to the nozzles of the spreading systems, INO maintains that its tool will make it possible to identify beetles and apply insecticides at the same time, strictly where the insects find. “What we are aiming for is a 25% reduction in the use of insecticides for the beetle. It would be major for the potato, ”says François Châteauneuf. The solution that INO is looking into seems all the more interesting to Philippe Parent, quality and agronomy director at Patates Dolbec, as it will also save him precious time in his fight against beetles. “A beetle is small. We could not take a satellite or a drone to detect it. It would take weeks to scan our fields, and it would already be too late to intervene and protect the harvest. »

What future for drones?

Field monitoring using drones equipped with a multispectral camera is gradually spreading among agricultural producers in Quebec. However, the small aircraft “does not necessarily offer a solution to everything,” warns Nicolas Deschamps, of the company Drones des Champs, whose offices are in Laval. First, because the autonomy of the device remains limited by the capacity of its battery. Second, because no chemicals are yet approved for drone spraying in Canada. According to Nicolas Deschamps, the future of phytosurveillance of fields, in particular large areas, goes through the satellite. “The satellites are going to be more and more precise. Today, they have image definitions of the order of a meter, but they will quickly go down in the coming decade to have definitions of the order of a centimeter, for example. It will meet 95% of the needs of agriculture, ”said the entrepreneur, who sees the future of the drone in targeted actions, such as limited stress zones in fields and on more rugged terrain, for example. “Once diseases are targeted by algorithms at specific locations, then drone spraying may be more economical than traditional spraying. »


Will LAMP ever come here?

LAMP (Loop-mediated isothermal amplification): an isothermal amplification technique, considered a robust method in terms of sensitivity, tolerance with inhibiting substances present in the real sample, and which allows the detection of the result with the naked eye.

LAMP technology raises the hope of certain players in the agricultural world for the early protection of diseases. For them, LAMP, which is not currently used in Quebec, would offer an advantageous alternative to the very expensive PCR tests carried out in the laboratory. Hervé Van der Heyden, researcher in plant pathology at Phytodata, believes these disproportionate hopes for the moment. “The advantage of LAMP is that it is not very expensive, which means that advisers or agronomists could have this type of technology in their office or their truck, if it is clean” , explains the researcher. Currently, Hervé Van der Heyden sees the interest of LAMP in the pre-diagnostic services it can offer. “A consultant who is not sure of his visual diagnosis in the field could use LAMP to improve his understanding of the situation”, illustrates the researcher who is interested in this technology for the early detection of diseases in strawberries.
Source: INRS ­


Claude Fortin, special collaboration

We would love to thank the writer of this write-up for this awesome material

New Surveillance Tools – Not A Panacea, But Technologies Are Getting Refined | Homeland


We have our social media profiles here and additional related pages here.https://www.ai-magazine.com/related-pages/