Smart Insect Monitoring System (SIMS)
AI-based Insect Monitoring
Pro AgroTech is focused on developing automated insect pest monitoring technologies ever since the advent of the 2G mobile communication networks 12 years ago.
At the initial stage, our effort is to develop an automatic insect counting device to replace conventional pheromone traps, the most commonly used trap for monitoring insect pest, mostly Lepidopteran pests such as Diamond Back Moth (DBM) Plutella xylostella and Oriental Leaf Worm (OLW) Spodoptera litura.
Conventional monitoring network consists of a number of traps located in the fields. In this device, male moths of the target species are attracted by the pheromone released from a lure, mimicking the smell of a female moth of the same species, and were trapped by a sticky board or a one way trap mechanism. The number of moths trapped could only be counted and recorded manually by technicians during regular field visits. The population density and dynamics of the target pests was then charted to provide information about the density and dynamics of the pest, which is vital for the control measures.
The development of an automatic counting device would make data collection a lot easier by relieving the burden of field trips. With more traps being installed, the data is more reliable. The challenge of automatic trap development is the accurate and reliable counting of live insect pests of various size, species, behaviour in the wild environment. Over the years, we have overcome numerous obstacles and developed a number of automatic insect traps.
-min.jpg)
Development of an array of pheromone-based automated trap sensors for insect pest monitoring (focus on various types of trap sensors)
Data transmission and processing are as important as data collecting. So we developed a data transmitting system based on telecommunication network through which data collected by the automated trap can be transmitted to a central platform where data is processed and analysed in combination with the biology of the target pests, the development of crops, and the meteorological data. This integrated system, consisting of a software platform, a number of electronic insect sensors with specific pheromone lures, significantly changed the way that insect pest data was collected, transmitted and processed, achieving an efficient and real-time insect monitoring system.This system has become a useful tool for plant protection agencies for directing control measures for insect pests of economic importance. It is also used by relevant authorities as an early warning system for invading or quarantine pests.
The automated insect pest monitoring system can also be used for monitoring insect pests in agriculture, orchards, and forestry.
Smart FAW monitoring in corn field
Trapped adult FAW moth
The same technology can also be used to monitor vector insects like mosquitoes for human diseases. A controlled release of CO2 venting mechanisms was used instead of pheromones for attracting mosquitoes.
The current system relies on specific pheromone lure for attracting the target insects to the sensor, limiting the use of the system to those species whose pheromone components have been well studied and a lure is available. For other insect pests of economic importance whose pheromone is not clear, UV light is usually used in the trap, creating a mixture of many insect species in the trap. A different technology would be required to identify the target insect species in this situation.
At Pro AgroTech, we aim to investigate the possibility of using Artificial Intelligence (AI) for insect recognition and identification. The goal is to develop an AI based product for insect recognition and species identification for those without pheromone lure.
Rodent body shape analysis - Length of the tail and the body, Contour extraction of the rodent pest, fur color, image binarization, gait analysis, synthetic body minutiae analysis
