Researchers have identified a novel way to detect stored product pests – using an electronic nose.
Detection of stored product pests is challenging, primarily due to their small size and cryptic habits – often hiding within stored foods away from light. This is a significant problem at all stages of the supply chain from mass grain storage facilities to food manufacturing plants to the pantries of residential homes.
Typically, current monitoring systems use pheromone traps to detect pests in the area. However, stored products still have to be inspected manually, which is a very time-consuming exercise requiring expertise. But it is not foolproof, as it is easy to miss infestations, especially for pests that hide away from the surfaces of the stored products or when there is only a low-level infestation. It is also an invasive technique.
The possibility of using image detection is being investigated. Whilst it has some benefits (automation and identification), it still doesn’t overcome the challenge of detecting pests that hide away from the surface of the stored products. With stored product pests and infested grains giving off specific odours, the possibility of using ‘electronic nose’ technology to detect stored product pests has been under investigation in recent years.
Electronic nose technology consists of a sensor which detects volatile organic compounds (VOCs) being released from the stored product. These VOCs are identified and then compared to reference profiles for successful identification. There are two learning models used by the software – using human-labelled data to train prediction models and ‘unsupervised’ learning, where AI is used to discover patterns in unlabelled data.
A number of trials have been carried out using different commercially available sensors over recent years. Most have been focused on a specific pest in a specific stored product and the trials have used different sensor technology and software. Nevertheless, accuracy in these studies is generally greater than 80%. Clearly, there is room for improvement, but results have been encouraging, with the ability to detect very low levels of infestation.
The electronic nose technology has two additional potential uses in stored product management. Mould and fungal growth are also a significant issue in stored grains. Grains affected by mould and fungus, which are hard to see, will also give off specific odours and so could readily be detected by an electronic nose. The technology can also be used to assess grain quality, which is a significant measure for crops – grains of different quality will have a different odour profile.
It may be a while before commercial detection equipment will be available for use in agriculture and pest management, but this technology has great potential.
However, key to having an electronic nose that can detect multiple pests is that a reference database of odour profiles needs to be developed – pest species, storage practices and storage conditions vary between regions and stored product, meaning a significant reference library is required for the electronic nose to accurately identify the issue at hand.
Further reading: Badgujar, Chetan & Swaminathan, Sai & Gerken, Alison. (2025). Electronic Nose for Agricultural Grain Pest Detection, Identification, and Monitoring: A Review. 10.48550/arXiv.2505.01301.
