Correctly detecting and identifying diseases and pests can save large proportions of crop yields every year. With artificial intelligence models, however, farmers no longer need to rely on expert human evaluations to perform accurate crop disease and pest identification.
These systems come in the form of phone-based apps or in drones and robots.
Crop disease and pest identification saves, on average, about a quarter of yearly crop yields. These AI systems, then, save livelihoods and contribute to local and global food security. Further, they do this faster than traditional methods.
Accurate identification systems also decrease the costs and deleterious effects of disease and pest misidentification.
An effective crop disease and pest identification system uses the farmer’s crop and property boundary data. This information ensures, first, that desired crops are protected and, second, helps identify local disease strains and pests. Additionally, this boundary data can ensure that remote identification systems do not trespass into neighboring farms.
These AI systems also include the farm’s crop disease and pest history, especially fertilizers and herbicides previously used and currently in stock. The identification system primarily uses this information to suggest pest and disease management solutions. However, this information also suggests alternate disease and pest identification. Some disease strains, for example, are chemical resistant, and the systems should tailor their recommended management solutions to this information.
Employee identification data should also be included so that unauthorized individuals cannot breach the farm’s database.
Good crop disease and pest identification systems incorporate a large amount of continuously updating external data on diseases, pests, and management solutions. Diseases and pests may be further divided into common and uncommon varieties.
Other important external data include news and weather history and forecasts. Weather, obviously, can make plants vulnerable to certain diseases. The news, meanwhile, can warn farmers and their AI systems that, for example, researchers identified a mutant disease strain impacting the crops they grow.
Additional external data may include other agricultural data, such as soil quality. A good system can use this kind of data measured by IoT sensors and incorporate it into an inclusive IoT farm ecosystem.
Industry data is another good resource, especially for farmers who may want to upgrade their crop disease and pest identification systems.
The main challenges of building and maintaining a strong crop disease and pest identification system include price, ease of use, and the need to constantly manage and update the system. Obviously, new plant diseases mutate all the time, and rodents and insects move into different parts of the world due to climate change, deforestation, and so on. Price and ease of use, however, are major roadblocks to adopting this technology, especially among subsistence farmers in many parts of the world. Making this technology available through mobile phone apps, however, can help these farmers.
Security is also a concern, especially for larger farms which use IoT-connected robots to track plant health.
ICRISAT: Artificial Intelligence to Track Pests and Diseases in India
NCBI: IoT-Based Strawberry Disease Prediction System for Smart Farming
Currently, the removal of infected trees is carried out after the visual confirmation of symptoms of HLB by trained inspectors teams. However, visual inspection takes a long time and is not sufficiently accurate. It is expensive and tedious.
VetorGEO’s research team has been developing a new detection method utilizing Agrowing’s high resolution 4 and 10 band multispectral sensors.